The body-discipline rubric cited git-commits as it stood before the ADR-0020
retrofit -- twelve Gotchas, 387/1102 words, line numbers :31-:52. Every
figure was correct for that version and none survives in the current file,
so the example is now anchored to commit 5e23250 and marked not to be
refreshed against HEAD.
More than staleness: row four called the secrets Gotcha a paraphrase FAIL
because step 2 restated it. Wave 2 followed that reasoning, deleted the
always-loaded copy, and left the amend branch able to commit a credential
unchecked -- dispatch loads exactly one flow file. The paraphrase rule now
carries its missing precondition: delete a restating Gotcha only when the
surviving copy is reachable from every branch that needs it, and relocate
a multi-branch safety gate into the body rather than dropping it.
Closes the second open design decision on PR #103. The `/compact` finding was
recorded as a false positive needing an allowlist or a suppression mechanism.
It is neither: the routing universe is the apm marketplace, so a target either
resolves to a skill or an agent or it does not resolve, and `/compact`,
`/clear` and `/init` are Claude Code slash commands with no counterpart in
Copilot CLI or Codex. `.apm/` source compiles for all three, so a
vendor-neutral description routing to one is a portability defect and the hard
FAIL is a true positive.
An allowlist was rejected for a concrete reason, not a stylistic one: it
answers a different question ("does this exist on some host?"), it cannot
answer that portably from a single source file, and it goes stale the next
time a host ships a command — reintroducing the same-commit-two-verdicts
failure ADR-0020 already closed for deployed trees.
Nothing is blocked today: zero of the 43 descriptions name a host built-in,
and an author who needs to mention one writes it un-slashed, which is not
route notation and carries no routing claim.
Recorded in ADR-0020 and in both author-facing contract references, so the
next agent reads the decision rather than "fixing" the gate.
ADR: 0020
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
A prose-form routing target blocks a commit only when its own sentence names
another target that resolves. That makes the sentence splitter part of the
ADR-0020 contract rather than an implementation detail, and the naive
"period, space, capital" rule got it wrong in both directions:
- OVER-SPLIT: `e.g. "..."` is not a sentence end, but the quote looks like a
start. The clause was cut in half and the corroborator stranded on the far
side, so a genuinely dangling target silently demoted to SUGGESTION — a
measurement taken and then discarded, the vacuous-green shape this gate
exists to prevent. Seven such splits are live in the current corpus.
- UNDER-SPLIT: a sentence opening with a code span or a lowercase skill name
was not seen as a start, so two sentences merged and a resolving target
vouched for an unresolvable one it never stood beside — a hard FAIL with no
escape hatch, which is the exact failure corroboration was added to prevent.
The splitter now excludes the five abbreviations that occur in routing prose
and admits a backtick or lowercase letter as a sentence opener. Applied
byte-identically to all three copies of the shared resolver.
Verified zero-delta against the corpus: 37 ERROR / 58 SUGGESTION / 2 dangling
before and after, findings byte-identical. The exposure this closes is to the
descriptions #99 is about to rewrite, not to the ones already measured — which
is why the deferral reason recorded on PR #103 ("can move the documented corpus
counts") does not hold and the fix lands here rather than after the retrofit.
Three regression tests, one per direction plus the backtick opener, each proven
non-vacuous by reverting the splitter alone and watching it go red.
Refs: #99
ADR: 0020
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Diffing each retrofitted SKILL.md against its replacement references/ files found
rules that existed on main and now existed nowhere — relocated in intent, deleted in
fact. A trim that loses a rule is not progressive disclosure, it is data loss with a
smaller word count.
Three had no survivor. The least-privilege guidance for `tools` kept its mechanics
and lost the "restrict to what the agent needs" half, so the remaining text read as
encouragement to omit the field. The improve flow lost its regression check, so
nothing compared the closing audit against the pre-edit state and a PASS quietly
becoming a SUGGESTION went unnoticed — restored on both halves of the author pair,
since agent-author had dropped its equivalent too. And agent bodies lost "would the
agent get this wrong without it?", which mattered more than it looks: ADR-0020
deliberately sets no body word gate for agents, three of the four already sit
between 933 and 1,199 words, and the delegation check only fires on procedure a
skill already owns. That heuristic was the only brake left.
Two more were reachable only from the wrong scope. agent-author tells the reader to
load only the file for the resolved scope, but the mcp__ glob syntax for
disallowedTools and the five tools no subagent ever receives had both landed in
project-user-scope.md. disallowedTools is the ONLY permitted fence at plugin/APM
scope, so the scope that needs the syntax most could not reach it, and a plugin-scope
run could write a body telling the agent to ask the user a question.
Two documents were actively wrong rather than merely thin. agent-audit told auditors
that validate.sh resolves boundary targets for skills only; it runs at both scopes,
so the auditor was hand-resolving what the script had already decided and could
contradict it. And skill-audit routed to its script-troubleshooting reference
whenever validate.sh "fails" — but it exits 1 on ordinary content FAILs, the normal
outcome for the whole #99 population, so 1,302 words loaded on nearly every audit.
A context-budget regression inside the skill that enforces the context budget.
Finally, two illustrations taught the shape the gate ERRORs on, unfenced, while an
adjacent rubric called it a hard ERROR.
LESSONS.md records the reference-chain depth rule flipping from "one level deep" to
"two hops, never three". ADR-0020 is silent on it and the reversal rode entirely on
the diff; the looser rule is what mandatory dispatch requires.
Refs: #99
ADR: 0020
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015W3iwF9ncfRZddGBxsMCYi
Review of the ADR-0020 gate found four ways it could exit 0 without measuring, and
one way it hard-failed a repo it had no business failing. On a gate shipping hot
with no baseline, a silent pass is the worst outcome available and a false block is
the second worst.
Consumer resolution was the blocker. _authoring_root() fell back to the nearest
.git, so it returned truthy in ANY git repo; _collect_authoring_root() then
contributed nothing and the deployed-tree branch was dead code in precisely the
consumer case it exists for. A consumer repo routing to an installed sibling got an
unblockable ERROR, and deleting .git "fixed" it. It now keys on which of the two
walk-up passes matched. A name-count delta was tried first and is wrong: a
single-plugin monorepo re-collects its own package and adds no new name, so the
delta reads zero and drags the deployed trees — including a global ~/.claude — back
into the universe. That reintroduces the install-dependence ADR-0020 forbids, one
layer down.
The three silent passes: an indented `---` inside a block scalar truncated the
frontmatter and reclassified the rest of the description as body; a non-string
description was str()-coerced, so `description: true` measured as the four-character
"True"; and an unterminated fence blanked the rest of the body, disabling the
ERROR-tier references/ check and the gotcha counts.
Two measurement defects came with them. The awk line/word counts discarded awk's
exit status, so an unreadable file passed both spec ceilings in total silence, and
awk NR/NF disagreed with the audit script's splitlines()/split() on Unicode
whitespace — the "fix one gate, get blocked by the other" bug, on the two axes the
differential test deliberately excluded. Both counts now run in the Python block
that already reads the file. A type error also no longer reports itself as a syntax
error.
Also: glob metacharacters in the checkout path silently disabled the resolver;
re.I was applied to some extraction patterns and not others; agent-audit missed
`tools:` written as a YAML block sequence, the shape Copilot files use; and a
nonexistent agent file raised a bare FileNotFoundError instead of a diagnostic.
The shared resolver block stays byte-identical across all three scripts. Corpus
output is unchanged — 26 description FAIL, 9 body FAIL, 2 dangling, 0 missing
references, 58 SUGGESTIONs — so no documented count moves.
Refs: #99
ADR: 0020
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015W3iwF9ncfRZddGBxsMCYi
Six defects, each one a place where two files that an author reads in the same
sitting told them different things — or where the trim dropped a rule and nothing
noticed because no gate covers prose.
**"Use proactively" contradicted itself across the pair.** All three agent
templates said to add it where the runtime should delegate unprompted, while
`agent-audit`'s `KyberforgeCopilot.ProactivePhrase` rule grades it a hard FAIL in
any `*.agent.md` — which is the Copilot half of every project/user pair *and* the
vendor-neutral plugin-scope file, since that compiles to a real Copilot agent
downstream. Following the template produced a file the repo's own gate rejects.
The phrase is now permitted in exactly one place, the Claude Code `.md`, and
`references/contract.md` carries the per-file table plus the consequence authors
ask about next: a pair whose CC half has it and whose Copilot half does not is
correct, because `agent-audit` checks that both halves describe the same job, not
that they match word for word.
**The output-schema rule contradicted itself inside one file.** `contract.md`
said any content only one branch reaches moves to `references/`, and then offered
an "Output format template" body pattern with no qualification. Stated once now,
so it is not re-litigated: an output schema stays in the body only when every flow
produces it and it is roughly 50 words or less. No third option.
**Gotchas tiers disagreed with the script.** `validate.sh` emits the entry count
through `suggest()` and exits 0, while `skill-author` and `skill-audit` both
called more than five entries a FAIL. Whether a given gotcha earns its place is
judgment, so the prose moves to the script's tier rather than the reverse. The
paraphrase rule stays a FAIL and is explicitly marked as the auditor's call — no
script detects it.
**The dispatch exemplar was cited at the wrong number.** `apm-workflow`'s body is
421 words; 554 is its whole-file count. Both `contract.md` and `body-discipline.md`
cited 554 while describing a body budget, so an author calibrating against the
exemplar overshot by ~30% — the exact whole-file/body-only conflation those two
sections exist to warn against, reproduced inside the warning.
**"Error handling" came back as a required body element.** It was one of four and
is the one that gets dropped, and dropping it is not neutral: an agent handed
malformed input with no instruction invents a recovery, and a subagent's invented
recovery is invisible to its caller until the output is wrong. Restored in
`agent-audit`'s rubric as a SUGGESTION, in `agent-author`'s contract and both
scope checklists as a required element, and as an `## Errors` section in all three
templates.
**`skill-author` Step 4 gains the one check the audit misses.** An empty body
reports `PASS SKILL.md body word count 0` — a word gate cannot tell "concise"
from "absent". Step 4 now hand-checks for a non-empty section, and its commit
verification is conditioned on actually being inside a git worktree, which a skill
under `~/.claude/skills/` is not.
Also here: absolute repo paths removed from `skill-author`'s SKILL.md and
contract.md in favour of naming the skill (`zoom-out`'s description is quoted
inline instead of pointed at), the boundary-target universe documented to match
the resolver, a two-hops-from-SKILL.md limit on reference chains, and
`new-agent.sh`'s next-steps output naming the description budget and the
deliberate absence of an agent body gate.
Refs: ADR-0020
ADR-0020 shipped its gates hot with no baseline file, so 26 of 39 descriptions
and 9 of 39 bodies are over their FAIL tier and editing any of them for any
reason requires bringing the skill into contract first. `references/improve.md`
said exactly that and stopped there — it mandated a retrofit and supplied no
procedure for one.
Four dry-run retrofits confirmed what that costs. Asked the same questions —
what to cut first, when a body is two flows rather than one, what else has to
change alongside — they invented six to ten different answers, so the same skill
retrofitted twice produced two different skills and neither run could be reviewed
against anything.
`references/retrofit.md` fixes the answers: an ordered cut list ranked by tokens
removed against behaviour lost (inverting that order is how a retrofit deletes the
instruction the skill existed to carry), the test for whether a body holds two
mutually exclusive flows, the reference-file conventions, the collateral checklist
for `README.md` and `references/sources.md`, and a worked description retrofit.
It also states the trap the dry runs kept hitting: retrofit the skill in place,
inside its package. The boundary-target universe is built by walking up from the
file being checked, so a scratch copy has no authoring root above it, the check
prints `INFO ... DID NOT RUN`, and the run still exits 0 — a line that reads as a
pass and is not one. A retrofit signed off on a copy carries an unverified
boundary target into the corpus.
Loaded from the improve flow only when a budget is actually exceeded, so a routine
improvement pays nothing for it.
Refs: ADR-0020, #99
The ADR-0020 body trim took `skill-audit` from 2,623 body words to a dispatch
shape, and two things went out with it that were not padding.
The manual structural fallback was one. Its replacement was a single sentence
telling the auditor to report an INFO when `validate.sh` cannot run — so with no
`python3` or no PyYAML, `skill-audit` reported the gap honestly and then audited
nothing structural at all. Every ADR-0020 measurement, the whole-file ceilings,
the name-to-directory match, the `references/` pointer check and the script
hygiene checks silently left the audit. A skill's whole Structure dimension
hanging on one optional interpreter is the same vacuous-pass shape the gate
scripts were just fixed for, one layer up.
The `E100 Runtime error ... does not exist` diagnostic was the other. That exit
code means an explicit relative `--config` was passed to `vale-wrap.sh` while
vale itself was installed and working; without the note, Step 1's fallback reads
exit 2 as "vale unavailable" and downgrades the description, body-discipline and
patterns dimensions to full LLM judgment for a config error it could have fixed.
That misreading is already recorded in CONTEXT.md as the reason both audit skills
stopped passing `--config` at all.
Both are restored in `references/validation-scripts.md`, loaded only when a Step 1
script fails — so the body pays nothing for them on a clean run, which is what the
dispatch pattern is for. The file also carries the by-hand boundary-target
procedure and the three ways to misread the result, including that
`INFO ... DID NOT RUN` is not a pass.
`references/file-structure.md` gains the one sanctioned spelling for a cross-skill
reference. The possessive form (``skill-audit's references/validation-scripts.md``)
is the only spelling both rules accept: a full repo path is what that section
already forbids, and a bare `references/<file>.md` is now a hard ERROR from the
ADR-0020 pointer check, which requires the file to exist in the skill's *own*
directory. Without the rule the two constraints look mutually exclusive.
Refs: ADR-0020
Three ways the gates could report green having measured nothing. All three were
invisible to a passing test suite, because pre-commit prints nothing at all for a
hook that exits 0 — a gate that declines to check and a gate that checked and
passed produce the identical signal.
- A UTF-8 BOM, a leading blank line, a trailing space after a `---` marker or
CRLF line endings defeated the `^---\n` frontmatter matcher. Every ADR-0020
check was then skipped and the file passed: measured at the time, a
550-character description with a 1,000-word body exited 0 behind a BOM.
All four shapes are now tolerated, and frontmatter that genuinely cannot be
parsed is a hard ERROR rather than a silent skip.
- An agent file with a valueless `description:` followed by another key let a
line regex capture the *next* key, which looked non-empty, so the
missing-or-empty branch never fired and every gate below it early-returned on
the empty folded value — zero output, exit 0, on a blocking gate. The one
field this contract is entirely about was the one field a gate could fail to
notice was absent. Presence is now decided on the YAML-folded value and
nowhere else, and a missing or empty description is a hard FAIL in all three
validators.
- The hand-rolled frontmatter fallback disagreed with PyYAML across the FAIL
boundary on folded scalars, so which reader happened to be available decided
the verdict. A fallback that mis-parses a scalar shape reports a vacuous pass,
which is worse than not running, so it is deleted: python3 and PyYAML are hard
requirements that fail loudly with an install pointer.
Boundary-target resolution no longer derives its universe from its own location.
A `${BASH_SOURCE}`-relative repo root leaked this repo's 39-skill universe into
every consumer repo running the hook through pre-commit, so a consumer skill
routing to `skill-audit` resolved against a plugin it had never installed. The
interim form resolved through `.claude/` and `.agents/`, which are gitignored
`apm install` output — the same commit reported 2 dangling targets on a machine
that had run the install and 6 on a fresh clone. Resolution now walks up from the
file being checked to an authoring root (nearest ancestor holding
`plugins/*/.apm/{skills,agents}`, else the nearest `.git`, in two passes so a
nested `.git` cannot outrank a real monorepo root); the universe is every skill
and agent under `<root>/plugins/*/` plus the file's own apm package and that
package's declared `dependencies.apm`. Deployed trees are consulted only when no
authoring root exists at all — the consumer case. One commit now gets one verdict,
which a gate shipping hot with no baseline file has to.
Narrowed in the same pass: a routing target inferred from the prose boundary form
and corroborated by nothing else reports at SUGGESTION instead of blocking. A
blocking check with no escape hatch is the wrong trade when the inference from
prose is the weak part of it.
New deterministic checks, all previously untested or absent: every
`references/<file>.md` a body names must exist (ERROR — a broken pointer is not a
style opinion); a description with no boundary clause at all, a Gotchas section
over five entries, and a Gotchas section over 25% of the body are SUGGESTIONs.
Where no universe can be determined the target check prints `INFO ... DID NOT
RUN` rather than passing quietly. Each prose-scanning check needed its own
false-positive fix — a fenced example of a Gotchas section was being read as the
section itself — and those fixes are pinned rather than assumed.
The resolver is one block copied verbatim into all three scripts between
BEGIN/END markers, because a cache-installed plugin's scripts cannot read outside
their own plugin directory. Nothing asserted the copies were still identical; a
one-line edit to a single copy passed every constant-agreement assertion, since
constants are not what drifts.
Tests land here rather than in a later commit. The existing suites assert the old
behaviour and go red against these scripts, so splitting them would leave a commit
whose own `run-tests` pre-push gate fails in isolation.
Refs: ADR-0020
Skill name+description pairs are preloaded into every session, costing
~6,200 tokens across 39 skills before any skill is invoked. The authoring
rules mandated that growth: skill-author:104 and description-quality.md:21
both required padding, while skill-author:102 (the deflating rule) had no
FAIL condition behind it.
Gates (blocking, no baseline file):
- description 250 chars SUGGESTION / 400 FAIL, measured on the folded
YAML value
- body-only 600 words SUGGESTION / 900 FAIL, independent of the unchanged
whole-file 2770-word / 500-line spec backstop
- every boundary-clause routing target must resolve to a real skill or
agent; catches skill-improve, neuledge-context and gitea-labels
- agents take the description gates but deliberately no body gate; a test
pins that absence
Vale: DescriptionOpener widened to ^This\b, new CompositionNote rule
banning architecture notes from descriptions. 10 hits, 0 false positives.
Kyberforge's own four skills retrofitted: descriptions 3,364 -> 938 chars
(-72%), bodies 8,306 -> 2,487 words (-70%), all via the apm-workflow
dispatch pattern. Fixes the skill-improve dangling route and the
agent-author misroute to manual review.
Also fixes a pre-existing false positive where any line-initial 'read '
was flagged as interactive input, which had already caused two scripts to
be rewritten around it.
Refs: ADR-0020
Output of scripts/sync-plugin-content.sh --all against this round's .apm/
source edits. No file here is hand-edited.
Carries agent-author's scaffolder and its contract doc into the flat mirror.
No compiled manifest changed: nothing in this round touched apm.yml, so apm
pack and sync-marketplace-mirror.sh both produced byte-identical output.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01X7GvKuJfy2WrdBmUttV4DT
Output of scripts/sync-plugin-content.sh --all against this round's .apm/
source edits. No file here is hand-edited.
Carries the agent-author and agent-audit documentation changes into the flat
mirrors. No compiled manifest changed: nothing in this round touched apm.yml,
so apm pack and sync-marketplace-mirror.sh both produced byte-identical output.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01X7GvKuJfy2WrdBmUttV4DT
Output of apm pack, sync-plugin-content.sh --all and
sync-marketplace-mirror.sh against this round's source changes. No file here
is hand-edited.
Carries the version bumps and marketplace owner.email into the compiled
manifests, the disallowedTools frontmatter and doc corrections into the flat
mirrors, and changes plugins/bin/.github/plugin/plugin.json's mcpServers from
the inlined server object to the ".mcp.json" pointer. That last file also
returns to 0644: the previous re-injection wrote it through mktemp and carried
0600 across, which no gate could see because the mode check did not cover
.github/plugin/ and git tracks only the exec bit.
.agents/plugins/marketplace.json is unchanged and that is correct -- apm's
codex profile carries neither version nor owner keys, so nothing in this round
reaches it.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01X7GvKuJfy2WrdBmUttV4DT
Documentation asserted things the code no longer did, which is how several of this
round's defects survived three review passes.
- `docs/spec/architecture.md` still described the pre-APM model — "each plugin has a
`plugin.json` manifest" — when no such file exists. AGENTS.md routes agents here
for the current structure, so it was live drift, not archive. Also cited ADR-0012
where ADR-0003 is meant.
- AGENTS.md never said `.apm/` is the sole hand-edited source, while the generated
mirror is byte-identical with no marker — an agent reading only AGENTS.md would
edit the mirror and find out at pre-push. It also omitted that `apm` and `jq` are
now required to push, unlike the `vale` bullet that exists for exactly that reason,
and understated the pre-push gate. `apm marketplace check` hard-fails offline and
`--offline` does not help, so the `SKIP=` escape hatch is documented rather than
new machinery built.
- ADR-0017 carried hardcoded line numbers that drifted twice within one PR, and said
`mcpServers` reinjection was real-syncs-only where the script's own header says
both modes. Line numbers dropped rather than corrected.
- ADR-0015 asserted issue #90 was closed. It is open. Reworded to defer to the
issue's own state rather than depend on someone remembering to close it.
- ADR-0014 said both audit skills pass `--config`; both SKILL.md files say pass none,
and passing one fails with E100. The ADR was wrong.
- ADR-0007 cross-referenced an archived ADR-0011 whose number was later reused.
Disambiguated with a note — renumbering a published ADR rewrites an immutable
record and breaks every existing citation.
- CONTEXT.md claimed `lint-runner` is report-only via a missing `Edit` tool.
Plugin-scope APM agents cannot express `tools:` (ADR-0016), so it is report-only by
instruction now; ADR-0016 accepted that regression but CONTEXT.md was not updated.
- Recorded two undocumented consequences of the conversion: `displayName` dropped
from all six compiled manifests and `owner.email` from the marketplace, and
`mattpocock-skills` silently version-pinned and manually maintained.
- Wrote the version-bump policy issue #90 specified but never delivered. Its previous
carrier, `marketplace-author`, was deleted in this PR, leaving the per-package rule
documented nowhere.
- README fixes: kyberforge listed skills belonging to the `git` plugin and a deleted
hooks path, `bin` advertised an agents directory it never had, and the docs index
listed files that do not exist.
Refs: #90
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01X7GvKuJfy2WrdBmUttV4DT
The mirror's `tests/` exclusion was depth-agnostic, so it deleted
`skill-author/assets/templates/tests/` — a template the skill scaffolds FROM —
alongside the depth-2 dev fixtures it was meant to drop. Since ADR-0017 makes the
mirror the installed content, the shipped scaffolder was broken: the mirror copy of
`new-skill.sh` exited 2 on `sed: can't read .../tests/README.md`, leaving a
half-written skill, while the byte-identical `.apm/` copy exited 0. `--check` was
green about it.
Check mode was restructured rather than patched because `diff -x` matches a basename
at any depth and cannot express the depth-2 scoping the fix needs — the two modes
could not be made to agree by construction. Check mode now runs the real `sync_dir`
into a throwaway root and diffs with no exclusions, leaving the exclusion rule and
the hooks destination each in exactly one place.
Also fixed here, all previously invisible to `--check`:
- Merged hooks were written to `<plugin>/hooks.json`, which Claude Code does not
convention-scan, while ADR-0017 itself quoted `hooks/hooks.json` as the contract.
Moved, with the legacy path cleaned up as stale. No `hooks` pointer is added to
`plugin.json`, so this does not reopen the option ADR-0017 rejected.
- Only the first drift per plugin was reported: `diff | sed` returns 1 under
`pipefail`, and `set -e` killed the subshell before the remaining checks and
before `FAIL=1`.
- File-mode and symlink drift were invisible, so `--check` and a real sync
disagreed; a find-based type/mode manifest now covers both.
The tests pinned almost none of this — the stale-skill wipe, the check-mode stale
branch, three `MIRROR_DIRS` entries and the hooks newline normalization could each
be deleted with the suite still green. All are now mutation-tested.
Refs: #90
ADR: 0017
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01X7GvKuJfy2WrdBmUttV4DT
Claude Code's (and Copilot's) native plugin installer has zero awareness of
.apm/ nesting -- it convention-scans only flat skills/, agents/, commands/,
hooks.json at each plugin's root. Confirmed via strings on the installed
claude binary and live installs of git@holocron/gitea@holocron/kyberforge@
holocron, all reporting Skills(0) Agents(0) Hooks(0) post ADR-0015's apm
conversion. Root cause (apm_cli/core/plugin_manifest.py): apm's plugin.json
compiler deliberately strips skills/agents/commands keys, assuming the host
already auto-discovers those convention directories -- it has no model of
.apm/ being host-visible at all. Separately, apm's own bundle exporter
(apm_cli/bundle/plugin_exporter.py, behind `apm pack --format plugin`)
implements the correct .apm/ -> flat mapping, but only ever targeted
build/<name>-<version>/, a path nothing in marketplace.json's source: points
at.
scripts/sync-plugin-content.sh wraps that bundle exporter and copies its
agents/, skills/, commands/, instructions/, extensions/, and merged
hooks.json back into each plugin's own root as a second tracked
compiled-output category -- same governance status as
.claude-plugin/plugin.json: generated from .apm/, never hand-edited. tests/
subdirectories are excluded from the mirror (dev fixtures, not host-visible
runtime content; several hardcode a relative repo-root walk-up sized for the
.apm/-nested depth, which breaks when duplicated one level shallower).
Applied for real across all 6 plugins and verified two ways: `claude plugin
validate --strict` passes on every real plugin directory, and a live
`claude --plugin-dir <path> -p "list skills/agents"` behavioral test
confirms content is now actually discovered.
Also, from the same issue #90 review round:
- scripts/check-manifests.sh pointed at each plugin's root-level plugin.json
(checking skills/hooks/mcpServers/agents pointer fields) -- that file was a
stale near-duplicate of .claude-plugin/plugin.json nothing else read or
wrote, now deleted across all 6 plugins. check-manifests.sh is rewritten to
validate .claude-plugin/plugin.json instead, and drops the pointer-field
checks entirely (nothing to check -- those fields are correctly absent by
design). Content-presence drift is now check-plugin-content-sync's job, a
new pre-push hook wired in .pre-commit-config.yaml.
docs/adr/0017 records the root cause and decision in full, including two
rejected alternatives (patching plugin.json's path fields directly -- apm's
compiler strips them on every run; pointing marketplace.json at apm pack's
build/ output -- a version-suffixed non-source directory nothing can install
from without an extra build step). ADR-0015 and CONTEXT.md are updated to
point at it.
Refs: #90
Why:
ADR-0015 established that Microsoft APM (apm.yml + .apm/) should replace
this repo's hand-authored plugin.json/marketplace.json model, with those
files becoming compiled output of `apm pack` instead of files edited by
hand via the (now-retired) plugin-author/marketplace-author skills.
Issue #90 was the deferred execution of that decision, gated on #88
(apm tooling) and #89 (apm-native agent-author/skill-author routing).
Implementation notes:
- All six plugins (bin, core, git, gitea, kyberforge, lint) now carry
apm.yml + .apm/{skills,agents,hooks} as their authoring source. Skills
moved with a plain git mv (content-identical across targets). Agents
were re-authored, not moved: per ADR-0016, .apm/agents/*.agent.md
compiles verbatim to both Claude and Copilot, so plugin-scope agents
now carry only name/description/model/source_keys -- no tools: field,
no Claude-only knobs (isolation, maxTurns, effort, memory,
permissionMode).
- Root apm.yml registers all 7 marketplace packages (6 local plus
mattpocock-skills as a remote entry) under versioning: per_package,
matching this repo's existing independent-plugin-versioning practice.
- .claude-plugin/marketplace.json and every plugin's plugin.json are now
apm-pack-compiled output, verified against the prior hand-maintained
content: same names/descriptions/versions/licenses/authors, only
cosmetic serialization differences (JSON key order, owner email vs.
url, Unicode escaping).
- plugin-author and marketplace-author are retired now that apm-based
authoring fully replaces their job; kyberforge bumped 1.3.1 -> 1.4.0
for that removal, and the root marketplace catalog bumped
0.3.1 -> 0.3.2 to match, per the version-bump convention now
documented in apm-workflow's reference docs instead of a dedicated
script (apm has no native version-bump automation).
- Fixed hardcoded pre-.apm/ path assumptions across
.pre-commit-config.yaml, .pre-commit-hooks.yaml,
scripts/check-scope-walkup-sync.sh, scripts/sync-vale-styles.sh,
scripts/check-vale-style-sync.sh, six plugins' root plugin.json
(stale skills/hooks/agents pointer fields that check-manifests.sh
validates), and several tests/*.bats and tests/*.sh fixtures --
including a bats REPO_ROOT relative-path depth bug (10 files, one
extra .apm/ directory level to walk up) and a vale probe-path
isolation regression introduced mid-fix.
- Corrected empirically-wrong assumptions surfaced this session in
apm-workflow/apm-install's own reference docs: `apm marketplace
package add` does not accept local paths (only owner/repo remote
shorthand -- local packages are registered by editing apm.yml's
marketplace.packages[] directly); `apm compile` is a consumer-side
AGENTS.md/CLAUDE.md generator, not the plugin.json producer, and
hard-fails on skill/agent-only packages without --clean; `apm plugin
init <name>` nests a stray subdirectory when run with a positional
name arg from inside a same-named directory; no native Copilot
marketplace output profile exists; .mcp.json is merged into the
compiled plugin.json content-aware and target-scoped, with no
dependencies.mcp entry needed for simple passthrough; pipx is the
correct pip fallback on externally-managed Python environments.
- Renamed agent-author's copilot.agent.md template asset to
copilot.agent.md.template so apm compile's recursive *.agent.md glob
stops misparsing the placeholder template as a real agent primitive.
Impact:
plugin.json and marketplace.json are compiled artifacts from here on --
editing them by hand is no longer the workflow; edit apm.yml/.apm/ and
run apm pack. CONTEXT.md's Plugin/Plugin marketplace glossary entries
reflect this. ADR-0001 is marked superseded, ADR-0006 moot, and
ADR-0010 updated for the new .apm/agents/ path (project/user scope
unaffected, per ADR-0016). Full local verification: claude plugin
validate --strict on all 6 plugins, apm audit --ci, apm marketplace
check, check-manifests.sh, and the full test suite (165/165 bats,
13/13 shell scripts) all pass clean.
Fixes: #90
Refs: #88, #89
ADR: 0015
ADR: 0016
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Ub96PyaSRD9BHPktotj1pC
field-inventory.md's apm-agent-allowlist and validate.sh's runtime
check already included source_keys as a 4th allowed field, and the
apm-agent.md template already instructed authors to add it for
provenance tracking — but SKILL.md (x2), README.md, ADR-0016, and
deployment-modes.md still described the allowlist as name/description/
model, "nothing else". The template itself even contradicted its own
source_keys guidance with a header claiming "ONLY the three fields
below — full stop" directly above it.
Updates all six locations to document source_keys as the intentional
4th field, resolving the contradiction.
validate.sh's detect_scope() and validate-provenance.sh's
find_plugin_root() disagreed with new-agent.sh's already-correct,
documented walk-up semantics on three points, each causing validate.sh
to false-FAIL a legitimately-scaffolded project-scope agent pair:
- a marker-less directory walked up into $HOME (no .git/apm.yml of its
own) was classified as user scope instead of project scope
- the .git-boundary branch returned the walked-to .git location instead
of the conventional scope root, breaking any <root> that is a
subdirectory of a larger git-tracked tree (monorepo package dirs)
- the new conventional-root arithmetic introduced to fix the above two
cases had no guard against non-conventional/hand-placed file paths,
which could point it at the wrong ancestor
Also adds scripts/check-scope-walkup-sync.sh, a behavioral drift-guard
(per ADR-0014's no-cross-skill-path precedent) that cross-checks the
four independently hand-ported walk-up implementations (validate.sh,
validate-provenance.sh, new-agent.sh, new-skill.sh) against real
fixture scaffolds, wired into .pre-commit-config.yaml at pre-push so
future drift between the ports is caught automatically.
Verified via bash tests/run-tests.sh (13/13) and targeted before/after
reproduction of each bug this closes.
A fresh /code-review of the APM-native authoring retarget (PR #93) found
several correctness bugs beyond the ones already fixed on this branch:
- new-agent.sh silently walked a marker-less subdirectory under $HOME up
to user scope, contradicting its own usage text ("user scope is checked
directly, no walk-up") and risking scaffolding into shared global
~/.claude or ~/.copilot directories instead of the intended local path.
- The hand-copied apm.yml type: manifest detector in new-agent.sh and
new-skill.sh accepted mismatched quotes (e.g. `type: "skill'`) that
validate.sh's regex correctly rejects, and silently dropped a final
apm.yml line lacking a trailing newline — causing the scaffolder and
validator to disagree on scope for identical input.
- Plugin-scope agent frontmatter could still contain the apm-agent.md
template's HTML comments at ship time with no audit signal, yet
apm compile copies frontmatter verbatim and <!-- --> breaks YAML
parsing on both downstream harnesses.
- ADR-0016 asserted agent-audit already implements a SUGGESTION heuristic
for tool-restriction-needing plugin-scope agents; it doesn't.
- agent-audit/README.md still described the old plugin-pair model this
PR replaced with a single-file allowlist model.
- validate.sh's project/user-scope CC-only/Copilot-only field checks and
counterpart-missing check lost their only test coverage when the old
plugin-pair fixture was deleted.
Also replaces an echo-into-sed two-value parse (4 forks per call) with a
single space-separated echo + read in both scaffolders.
Regression tests added for every fix above, including one for a bug this
pass introduced and the test suite caught: an initial two-line
echo + `read` attempt silently dropped the second value, since `read`
consumes only one line regardless of embedded newlines.
Full suite: 158 bats tests, 39 shell-script tests, 12/12 summary
categories, 0 failures.
Refs: #89, #93
Post-implementation review of PR #93 (issue #89's apm.yml-native retargeting
of skill-author/skill-audit/agent-author/agent-audit) found four confirmed
defects across the four scripts' apm.yml `type:` walk-up logic:
- field-inventory.md's apm-agent-allowlist was missing `source_keys`,
contradicting agent-author/SKILL.md's own instruction (Step 5 checklist)
to allow it at plugin/APM scope — a correctly-authored file with
source_keys failed validate.sh.
- validate.sh's APM_TYPE_RE and validate-provenance.sh's TYPE_RE disagreed:
the former tolerated a quoted `type: "skill"` value, the latter didn't,
despite agent-audit/SKILL.md explicitly documenting that
validate-provenance.sh walks up "the same way validate.sh does". Both
also used `\b` word-boundary matching, which false-matches a malformed
value like `type: prompts-only` on the `prompts` prefix. Unified both
regexes to be quote-tolerant and require an exact value.
- All four scripts' `.git` project-boundary check used isdir()/[[ -d ]],
which misses git worktrees where `.git` is a regular file (`gitdir: ...`)
rather than a directory. Switched to exists()/[[ -e ]].
- new-agent.sh and new-skill.sh had the same quote-intolerance as above via
inline `grep -qE` calls (new-skill.sh's also had the `\b` false-match
bug); replaced both with a shared-shape `is_apm_package_manifest` bash
helper matching the Python regex's semantics.
Four other findings from the same review turned out not to be bugs: a
bare `plugin.json` no longer signaling plugin scope is documented,
intentional behavior (agent-audit/SKILL.md:30, agent-author/SKILL.md:87),
deferred to issue #90's real plugin.json-to-apm.yml conversion — not
something this fix should reverse.
Verified via direct reproduction of each defect plus the full test suite:
147/147 bats tests, 39/39 shell-script tests, 12/12 summary categories.
Refs: #89
detect_scope() had the same bug class fixed in new-agent.sh (099bdec):
it checked for a .git directory before checking whether it had reached
$HOME, so a dotfiles-managed home directory (yadm, chezmoi bare-repo,
etc.) made validate.sh misresolve to project scope, deriving the
counterpart as ~/.github/agents/<name>.agent.md instead of the correct
~/.copilot/agents/<name>.agent.md and failing with a false "counterpart
file not found". Check the $HOME boundary before the .git check, same
fix shape as 099bdec.
Found via post-implementation review of issue #89.
check_file()'s is_plugin_scope param and its plugin-silently-ignored
field check were unreachable dead code left over from the issue #89
restructure: plugin/APM scope now exits via check_apm_agent_file()
before check_file() is ever called, so is_plugin was always False.
Remove the param, its branch, the unused plugin_ignored_fields parse,
the now-stale field-inventory.md section, and the SKILL.md mention.
Found via post-implementation review of issue #89.
new-agent.sh's walk-up checked for a .git directory before checking
whether it had reached $HOME, so a dotfiles-managed home directory
(e.g. `~/.git` from yadm or a bare-repo checkout) made `new-agent.sh
<name> ~` silently resolve to project scope instead of user scope,
writing .github/agents/ instead of ~/.copilot/agents/. Check the
$HOME boundary before the .git check so it can no longer be shadowed.
Found via post-implementation review of issue #89.
skill-audit's cross-plugin path check now also recognizes .apm/skills/
<other-skill>/ as a breaking reference shape, alongside the existing
plugins/<plugin>/skills/<other-skill>/ pattern — the APM-native equivalent
of the same cache-isolation problem.
forge.md's Step 4 plugin-version-bump detection switches from "any
directory containing a plugin.json" to the same apm.yml+type: walk-up
used elsewhere in this batch, and hands off to apm-workflow (bump the
single apm.yml version) instead of the deprecated plugin-author.
Refs: #89
Validates the new single-file .apm/agents/<name>.agent.md shape agent-author
now produces at plugin/APM scope: frontmatter allowlist (name/description/
model only, from a new apm-agent-allowlist entry in field-inventory.md),
no counterpart derivation, and Pair Consistency dropped from that scope's
report entirely (nothing to pair by design). Adds a plugin/APM-scope-only
SUGGESTION when an agent's description/body implies a tool restriction or
Claude-only behavior the vendor-neutral frontmatter can no longer express
(ADR-0016).
Scope detection in both validate.sh and validate-provenance.sh switches
from a flat plugin.json/.claude-plugin/plugin.json check to a walk-up for
the nearest ancestor apm.yml with a top-level type: field, skipping
type:-less marketplace-only manifests — full switch, no dual-mode fallback
to the old plugin.json signal. validate-provenance.sh's walk-up was fixed
to match validate.sh's (it still used the old plugin.json check, and its
counterpart-merge logic was rewritten to read a single file's source_keys
instead of merging a CC+Copilot pair, since plugin/APM scope has no
counterpart). Project/user scope validation is unchanged in both scripts.
Refs: #89
Plugin scope now authors a single vendor-neutral .apm/agents/<name>.agent.md
file (name/description/model/body only) instead of a Claude Code + Copilot
CLI pair, per ADR-0016: apm compile has no per-target field integrator, so
tools: and all Claude-only fields (isolation/maxTurns/effort/memory/
permissionMode) are omitted entirely rather than shipping a value that's
guaranteed wrong on one harness. Scope detection switches from a flat
plugin.json check to the same apm.yml+type: walk-up agent-audit's
validate.sh already uses. Project scope and user scope are unaffected —
both keep the existing dual-file pair model.
Refs: #89
Skills now migrate from hand-authored plugin.json to Microsoft APM
(apm.yml + .apm/) per ADR-0015/issue #89. skill-author's write location
switches accordingly: walk up from the destination to the nearest
ancestor apm.yml declaring a type: field (skipping type:-less
marketplace-only manifests) and scaffold into <package-root>/.apm/skills/
<name>/. Standalone authoring (no apm.yml/.git anywhere, e.g.
~/.agents/skills/) is unchanged — only the old plugin.json-based path is
replaced, not standalone mode.
Refs: #89
Two review passes on PR #91 (correctness + simplification) found the
prior fixup commit (d6fd9b6) left the deprecation notice it added
half-applied, cited a nonexistent ADR, and introduced/compounded a
few duplicate restatements of the same facts across apm-workflow's
skill files and apm-orchestrate's Hard rules.
- plugin-author/SKILL.md cited ADR-0016 (doesn't exist) instead of
ADR-0006 for the version-parity invariant.
- The deprecation banner on plugin-author/marketplace-author wasn't
reflected in their routing `description:` frontmatter, forge's
routing table, either kyberforge README skill table, or CONTEXT.md
— now propagated consistently; routing behavior is unchanged since
issue #90 (repo-wide apm conversion) hasn't landed yet.
- apm-orchestrate's `configure` operation only covered scaffolding a
new package, not editing an existing apm.yml (e.g. removing a
marketplace.packages[] entry, which has no dedicated `apm` CLI verb
and is just a manifest edit). Added `edit-config`, intentionally
ungated — a local file edit isn't the same class of irreversible
action as `apm publish`, and normal review already gates the commit.
- Deduplicated facts restated 3-4x across SKILL.md/references/ADR
boundaries (audit semantics, reserved name prefixes, marketplace-add
direction, registries precondition), and trimmed configure.md's
~55-line apm.yml schema dump — copied near-verbatim from the
research doc — down to a compact key list with a pointer, matching
install.md's existing terse cross-referencing style.
Deliberately left as-is: apm-orchestrate's parallel/sequential
fan-out logic stays inline rather than extracted to a shared
reference, since neither git-orchestrate nor gitea-orchestrate need
it yet — premature extraction for a single caller.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0186ERbyACLRuRxPRnqwpa4m
Addresses 7 verified findings from the /code-review pass on PR #91's
APM-conversion changes:
- apm-workflow's description omitted install triggers, misrouting
"install my apm dependencies" requests to apm-install instead
- `apm marketplace check` was listed under both the marketplace and
audit dispatch rows, breaking the "one reference file per action"
contract; kept under marketplace.md, its real usage-sequence home
- ADR-0015 claimed apm-workflow docs are generic/repo-agnostic while
marketplace.md cited this repo's AGENTS.md and git-commits/
git-remotes skills by name; stripped the holocron-specific
citations (marketplace.md, compile.md) so the claim now holds
- plugin-author/marketplace-author are superseded per ADR-0015 with
deletion deferred to #90; added deprecation notices pointing to
the replacing apm-workflow dispatch actions
- apm-orchestrate/gitea-orchestrate/git-orchestrate all instruct
dispatch "via Skill" without granting the Skill tool; added it to
all three
- apm-orchestrate mandated strictly sequential fan-out across
independent packages with no shared state; relaxed to permit
parallel dispatch, matching this repo's own subagent-parallelization
guidance
Refs: #91
A grill-me session determined plugin-author/marketplace-author are fully
superseded by the landed apm-install/apm-workflow/apm-orchestrate artifacts
and will be deleted (not adapted) under issue #90, narrowing issue #89 to
skill-author/agent-author only; forge stays untouched. ADR-0015 is amended
to record this correction plus a three-way content-migration audit.
Claude Code platform-constraint facts that lived only in
plugin-author/marketplace-author (reserved name prefixes, the ADR-0010
agents/ stray-.md validator gotcha, claude plugin validate as a terminal
check, and a removal/re-pack confirm-before-proceeding note) remain true
post-conversion and are migrated into apm-workflow's reference docs via
kyberforge:skill-author, which also updated the skill's README and
sources.md provenance records. Dual-manifest-specific conventions (ADR-0006
parity, CC/Copilot field-placement split) do not carry forward since they
no longer apply.
ADR: 0015
Re-review (comment 24) of fix commit e16c3dc found six new issues, mostly
introduced by that fix commit itself: a dangling reference to a Hard Rule
bullet the same commit deleted (apm-orchestrate.md/.agent.md Process step
2 still named "secret indirection"), and an ADR-0015 Decision bullet that
claimed "this ADR does not update CONTEXT.md" while the same commit had
just added a forward-pointer sentence to CONTEXT.md's Plugin/Plugin
marketplace entries. Both reworded to match what actually happened.
apm-install's APM_INSTALL_DIR escape-hatch example dropped the curl pipe
entirely (`APM_INSTALL_DIR=... sh` with nothing piped into it) — fixed in
both apm-install/SKILL.md and the installation.md research doc, verified
against the upstream Microsoft APM docs via Context7.
Neither apm-workflow nor apm-orchestrate routed to plain `apm install
[PACKAGE_REF]`, the CLI command that actually resolves/fetches
dependencies declared in apm.yml — apm-install only bootstraps the apm
binary/runtime, not per-package deps. Added a 5th "install" dispatch
action to apm-workflow (new references/install.md, SKILL.md table row,
README usage/files sync, sources.md provenance entry) and a matching
"install" operation group on apm-orchestrate so it can route there.
configure.md's apm.yml schema block was also missing the "legacy singular
`target:` CSV form is still accepted" caveat its sibling research doc
documents for the same field — added for consistency.
The sixth finding (paired .md/.agent.md Output-contract disagreement) was
checked against git-orchestrate and gitea-orchestrate's existing pairs and
found to match established repo convention (JSON schema in .md, prose
summary without the enum in .agent.md) — left unchanged as a false
positive rather than "fixed."
kyberforge bumped 1.3.0 -> 1.3.1 via agent-author's normal improve flow.
Refs: #91
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0186ERbyACLRuRxPRnqwpa4m
ADR-0015's Decision bullet stated ADR-0001 was superseded as present
fact while Consequences framed it as conditional on issue #90 —
reworded Decision to match, and added a Supersedes header (repo
convention per ADR-0007/ADR-0010) so the pending change is visible
from the title. Also fixed a forward-note that misattributed the
pre-existing .agent.md convention (ADR-0005/0010) to the APM move,
and reworded the rejected-plugins/apm/ rationale to acknowledge the
shipped skills are actually generic rather than repo-specific.
CONTEXT.md's Plugin/Plugin-marketplace glossary entries got a
forward-pointer to ADR-0015 so a session-start read surfaces the
pending manifest-authoring change.
apm-orchestrate.md/.agent.md: removed a Hard Rule requiring ${VAR}
secret indirection that can never fire (the orchestrator has no
Edit/Write tool and never touches apm.yml — apm-workflow's SKILL.md
already owns this rule where content is actually written); removed
compile-manifest-check, an operation with no backing apm CLI command
anywhere in the research docs; added explicit CLI mappings for
add-package/add-marketplace so an executing agent can't invert them.
marketplace.md's release steps shelled out to raw git add/commit/
tag/push — replaced with a pointer to the git-commits/git-remotes
skills per AGENTS.md's "prefer plugin skills over raw shell" rule.
compile.md now states audit runs after compile/pack (audit scans
deployed output, not the source apm.yml) — answers a reviewer
question that had no documented answer.
Refs: #91
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0186ERbyACLRuRxPRnqwpa4m
Human-facing dispatch over apm's configure/marketplace/compile/audit
lifecycle, one reference file per concern, gitea-issues-style
dispatch table. apm-install handles the one-time binary/runtime
bootstrap that precedes this loop.
Two efficiency findings from a code-review pass:
- The separated (--config X) and joined (--config=X) argument branches
duplicated ~20 lines of absolutize-if-relative path logic. Extracted
into abs_config_value(), used by both branches; the redundant
--config=/* special case falls out since the helper already passes
absolute paths through unchanged.
- The common single-file path spawned python3 twice per file (once for
abspath resolution, once for flatten()). flatten() now optionally
takes a tmpdir arg and does both in one process. The per-file loop
under a directory argument is unchanged — that path wasn't flagged.
agent-audit's copy is canonical; skill-audit's copy was regenerated via
scripts/sync-vale-styles.sh, not hand-edited, to guarantee byte parity.
No hardening (bash 3.2 compat, surrogateescape, symlink guards) touched.
Verified: tests/test-vale-wrap.sh 39/39, check-vale-style-sync.sh clean,
full suite 12/12.
Both READMEs' file tables predate #85's Vale wiring and never picked
up scripts/vale-wrap.sh or the assets/vale/ style tree, so a reader
of either README had no way to find where the new Step 1 sub-check
actually lives. List the new files and note the Vale sub-check in
"What it does" for both skills.
skill-audit's Description dimension implied Vale's DescriptionOpener
alert fully covers imperative-phrasing checks, but that rule only
matches the literal "This skill..." pattern. agent-audit kept its
equivalent manual "does the description open with a verb" fallback
bullet; skill-audit's got dropped when Vale wiring landed in #85,
leaving other non-imperative openers (gerunds, passive phrasing) to
sail through unflagged. Restore the parallel check.
MAX_WORDS=2900 was calibrated to the corpus median density and carried no
margin: at the densest observed 7.22 chars/word (~1.81 tokens/word) it permits
~5,240 tokens against the 5,000 it proxies for. 2770 holds the worst observed
density under the ceiling. The largest SKILL.md is 2,489 words, so the change
costs nothing today — 281 words of margin — and the header comment now argues
the new calibration rather than swapping the digits.
Both enforcement points move together, and a new test asserts they agree, since
a SKILL.md passing its own audit while the commit hook blocks it is the
disagreement this pair exists to prevent.
CONTEXT.md is deliberately left ungated: it is 2,816 words, and gating it would
block the build. Recorded here so the omission reads as a decision rather than
an oversight.
skill-audit's manual-fallback path listed only the line ceiling, so an agent
taking that path passed an oversized SKILL.md the hook then rejected. The word
ceiling is now named alongside it. agent-audit is deliberately unchanged: the
size hook scopes to SKILL.md only and agent-audit's validate.sh has no word
gate, so claiming it there would be false.
The Vale research doc still showed the MDX {/* vale off */} form under a
Markdown heading, contradicting CONTEXT.md and vale-run's troubleshooting
reference — that form suppresses nothing in plain .md. Fixed in both places it
appeared.
tests/run-tests.sh used mapfile (bash 4.0+) with unguarded array expansion,
though AGENTS.md tells contributors to run it and macOS ships bash 3.2. It now
collects via a while-read loop over process substitution and guards every
expansion. The newline-delimited find|sort pipeline is kept rather than -print0
with sort -z, whose BSD portability is the weaker link, and which matches
mapfile -t's previous behaviour exactly.
Refs: #85
ADR: 0013
The flattener's last-resort branch rewrote ASCII ' to U+2019, justified as the
one combination no YAML scalar can carry verbatim. That claim was false: a |-
literal block with a single indented content line carries ', ", \ and ": "
verbatim and keeps text.frontmatter.description matching — as the wrapper's own
docstring already said of literal blocks. The rewrite fired on 12 of 54
in-scope files, silently disabling every rule whose token contains an
apostrophe. Case 20 pinned only that the scope stayed alive, so it passed
either way.
The emission site now splits the emitted scalar on its first newline so a
carried-over trailing comment stays on the "description: |-" header line rather
than becoming part of the value, and pads by span_lines - 1 - newlines. The pad
stays non-negative because the branch is only reachable when the original span
is at least two lines. Verified across all 73 in-scope files: no line-count
changes, and exactly the 12 expected files take the new branch.
One reported position moves: an alert on a description that is itself flagged
shifts from the key line to the block's content line, both inside the original
span. YAML cannot put a literal block's content on the key's own line, so this
is unavoidable; no line at or after the end of any description span moves.
Also: --output no longer absolutises the built-in style names line, JSON and
CLI, which a same-named file or directory in cwd turned into a template path
(exit 2, E100 Runtime error). And case 19's empty-baseline guard no longer
lets five dependent comparisons print vacuous passes — while fixing it the
guard turned out to be unreachable, since under pipefail an alert-free report
aborted the script at the assignment.
Refs: #85
ADR: 0014
Three divergences between what the audit skills claim and what the hooks
enforce, each of which fails silently rather than loudly:
- `skill-size-check.sh` blocked at 2900 words while `validate.sh` checked only
the 500-line ceiling, so `/skill-audit` could report a skill ready to ship
that the commit hook then rejected. `validate.sh` now checks the same pair on
the same inclusive terms; the constants are duplicated with a comment naming
the other file, because a plugin skill's scripts cannot read outside the
plugin directory once installed to the cache.
- Both audit skills' Step 1 passed `--config assets/vale/.vale.ini`, which is
redundant (the wrapper self-locates its sibling config) and fragile: an agent
that resolves the script path against the skill directory but not the config
path gets E100, exit 2, which the surrounding fallback clause misreads as
"vale unavailable" and downgrades to full LLM judgment with no signal.
- The external-consumer test registered only the two Vale hooks, never the
third shipped hook, so a lost executable bit would have broken every consumer
while the local suite stayed green. Verified by mutation: `chmod 644` on the
copied script now turns three passes into two failures.
Also corrects the size hook's calibration comment, which claimed ~5.7-6.5
characters per word against a corpus whose measured median is 6.79 — the stated
upper bound sat below the median, so the "calibrated with margin" claim was
inverted for prose-dense files. MAX_WORDS is unchanged pending a decision; the
comment is now explicit that the gate holds under 5,000 tokens for typical
prose density, not for any file.
Refs: #85
Vale locates a frontmatter description by matching the parsed YAML value back
against the source text, so any scalar whose value is not spelled out verbatim
loses the `text.frontmatter.description` scope entirely. The wrapper only
flattened `>` folded scalars, so plain, double-quoted and single-quoted
multi-line descriptions silently reported zero alerts and exit 0 — a clean pass
indistinguishable from a real one, in a prefilter whose callers are instructed
not to re-derive its verdict by judgment.
Implementation notes:
- Classify the scalar kind after `^description:[ \t]*` and reuse one shared
continuation-line generator for every form; `|` literal blocks keep their
line breaks, stay verbatim-matchable, and are still left untouched.
- Emit the flattened value in whichever scalar form needs no escape at all
(plain, then single-quoted, then double-quoted), because any escape breaks
the verbatim match. The old blanket `'` -> U+2019 substitution silently made
apostrophe-bearing rule tokens unmatchable across 63% of the corpus; it now
survives only for the one combination no YAML scalar can carry verbatim.
- Terminate continuations at a line flush with the key, not only on a shallower
indent — a `description:` followed by a flush-left line previously swallowed
the rest of the frontmatter.
- Route vale's value-taking flags explicitly instead of inferring targets by
file existence, and absolutize relative `--output`/`--path` values the way
`--config` already was, since the run `cd`s into the scratch mirror.
- Fail loudly on a nonexistent path instead of inheriting bare vale's fallback
to stdin, which rendered a typo'd path as `0 errors ... in stdin`, exit 0 —
a form the callers' `0 files` NOT-RUN guard cannot match.
- Follow symlinks when walking a directory argument, matching bare vale.
Refs: #85
Under set -u, "${arr[@]}" on an empty array aborts on bash before 4.4,
which is what macOS ships as /bin/bash. Three expansion sites now use
${arr[@]+"${arr[@]}"} consistently.
The hazard is not currently reachable: verified on a bash 3.2.57 built
from source that all seven invocation shapes succeed against the
previous code, including zero args, flags-only and an empty directory.
vale_args is provably non-empty at every site because the default
--config branch always appends first. The guard is kept because that
invariant is non-local and untested, so an edit to the default-config
branch would reintroduce a macOS-only crash silently.
Test fidelity is deliberately mixed. Case 16 is static and is the only
one that fails against the previous code, since no bash 5 host can
reproduce the abort at runtime. Case 17 runs the emptiest invocations
under the oldest bash it can find and names that shell in its output
so it cannot overclaim. Case 18 guards against the tempting wrong fix
of dropping the quotes, which also silences the abort but word-splits
a path containing a space.
No other bash 4.x construct is present; swept for mapfile, declare -A,
case modification, negative indices, globstar, wait -n and namerefs.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MCQ648fLSFXPHGZdQ8gn58
pre-commit prefixes only entry[0] with the hook-repo clone path
(cmd = (prefix.path(cmd[0]), *cmd[1:])), so the --config argument in
.pre-commit-hooks.yaml resolved against the *consuming* repo's root
and hard-failed every external run with E100. Two of the three hooks
ADR-0014 promises were unusable.
vale-wrap.sh now self-locates its config from BASH_SOURCE when no
--config is supplied; an explicit --config still wins in all three
argv forms and stays cwd-relative. Both manifests drop the argument
and are kept byte-identical: the local repo: local config resolved
--config correctly only because the consuming repo *was* this repo,
and that divergence is why three review rounds missed the defect.
Also in the wrapper:
- replace GNU-only `realpath -m` with a portable abspath helper; -m is
load-bearing (dest does not exist yet), so BSD realpath aborted the
script under set -e on macOS
- walk directory arguments instead of passing them through unflattened,
which reported a clean 0-error run for files that fail when named
explicitly
- read/write with errors='surrogateescape' so one non-UTF-8 .md under a
directory argument cannot abort the hook
New test-vale-hooks-consumer.sh builds the hook repo from the working
tree and points a file:// consumer at it, covering the manifest as a
hook repo for the first time.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MCQ648fLSFXPHGZdQ8gn58
skill-audit/agent-audit's Step 1 resolved vale-wrap.sh/.vale.ini via
`git rev-parse --show-toplevel`, which returns whichever repo the skill
happens to run in. Inside ai-development that works; in any external repo
that installs kyberforge@holocron as a plugin, it resolves to that repo's
own root, which has no .vale.ini — the prefilter silently fell back to
full LLM judgment. ADR-0013 named this as a deliberately deferred gap.
Vale's config/styles/wrapper now ship inside the plugin itself: a
canonical copy in agent-audit/assets/vale/ (Kyberforge + KyberforgeCopilot,
the superset agent-audit needs) and a smaller duplicate in
skill-audit/assets/vale/ (Kyberforge only) — per the no-cross-skill-path
rule already established for plugin cache-installs. Both skills resolve
these relative to their own directory, same as scripts/validate.sh
already does.
A new root .pre-commit-hooks.yaml exposes both copies plus
skill-size-check so any external repo can enforce the same rules via
`repo: <this-repo-url>, rev: <tag>` in its own pre-commit config,
independent of Claude Code entirely — the same mechanism covers CI. This
repo's own pre-commit hook now consumes the identical plugin-bundled
copies via repo: local (not a third root copy, and not a pinned
self-reference, which would lint working-tree edits against the last
tagged release instead of the change being made). Split into
vale-audit-prefilter-skill/-agent hooks after confirming, by diffing the
full corpus against both old and new config before deleting the old
files, that one combined hook pointed at only one copy silently 0-file-
skips the other file type.
scripts/check-vale-style-sync.sh guards the two copies against drift,
wired at pre-push alongside check-manifests.
ADR: 0014
Round-3 review of PR #85 found the "enforcing" pre-commit hook enforced
nothing. Vale's exit code keys on error-level alerts alone: five of the
six rules were level: warning, so they exited 0, and pre-commit hides
output from a passing hook — the alerts were invisible and blocked
nothing. ADR-0013 rejected a report-only trial tier and then shipped one
by accident.
Flatten every rule to level: error. Vale's own exit code is then correct,
so the hook entry drops to a bare vale-wrap.sh call and the graded
error->FAIL / warning->SUGGESTION mapping disappears from both audit
skills: every alert is a FAIL, in the gate and the audit alike. No
ignorable tier, matching shellcheck, the test suite and
conventional-pre-commit.
Delete Kyberforge.VagueQualifier. Measured against the 41 skill/agent
files as they stood before the rule ever ran: 2 hits. One marginal
("very different" -> "fundamentally different"), one an unfixable false
positive — caveman/SKILL.md quotes "of course" as an example of filler,
a mention not a use — which forced the only Vale suppression comments in
the repo. Those four lines go with it; two of them were dead anyway,
suppressing a frontmatter-scoped rule on a body line. Held-out prose (273
files) fired 15 times, 9 inside out-of-scope research examples and the
rest one word in two idioms in a single doc. SentenceOpenerThereIs
survives: 22 held-out hits, both in-corpus hits clean rewrites, zero
suppressions.
Widen .vale.ini's globs to [**/SKILL.md], [**/agents/*.md] and
[**/*.agent.md]. The plugins/*/-prefixed globs scoped nothing — Vale's *
crosses /, so they already matched docs/research/examples/**/agents/*.md
and assets/templates/SKILL.md, the two paths CONTEXT.md claimed they
excluded. Scoping is and was the hook's files: regex. The old globs also
hid a silent false negative: a skill outside plugins/ matched no section,
so Vale reported 0 files and exited 0, which both audits read as clean.
They now treat a 0-file run as NOT RUN and fall back to full judgment.
Also:
- vale-wrap.sh resolves relative --config values and file arguments
against the caller's cwd, as vale does, instead of the repo root, which
hard-errored from a subdirectory and silently skipped flattening for
file args that did not resolve from the root. Absolute paths inside the
cwd are relativized so reports cite resolvable paths, not scratch ones.
- vale-run's exit-code model was documented backwards ("exits non-zero
whenever it finds an alert at or above MinAlertLevel") and would have
led anyone following it to build a gate that passes everything. Its
Markdown suppression syntax was MDX-only and does not suppress in .md;
corrected in the skill and its troubleshooting reference, with
backtick/fence exemption documented as the first resort.
- skill-size-check.sh fails only above 500 lines, agreeing with
skill-audit's validate.sh <= 500 pass.
- ADR-0013 and CONTEXT.md amended to match, recording why graded
severities cannot gate.
Verified: 9 test scripts / 15 vale-wrap cases pass; vale-audit-prefilter,
skill-size-check and shellcheck pass --all-files; check-manifests and
claude plugin validate --strict clean. New tests fail against the old
script (3 of them) and pass against the new one.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MCQ648fLSFXPHGZdQ8gn58
Addresses PR #85's outstanding review items after grilling the open
questions against ADR-0013/CONTEXT.md/ADR-0010:
Blocking fixes:
- vale-wrap.sh: replace json.dumps() escaping (which silently defeated
Vale's frontmatter scope on any description containing a quote,
backslash, or non-ASCII char — ~58% of the corpus) with a single-quoted
YAML scalar, substituting a Unicode right single quote for embedded
apostrophes rather than '' doubling (Vale's frontmatter scanner isn't a
full YAML parser and silently truncates on '' too).
- vale-wrap.sh: fix a blank-line-inside-a-folded-description truncation
bug via indentation-based, blank-line-tolerant body capture; narrow
flattening to `>`-style scalars only (`|` already works unflattened).
- skill-audit/agent-audit Step 1: make the vale-wrap.sh invocation
cwd-independent via git rev-parse --show-toplevel, fixing a bug where
no single cwd satisfied all three Step 1 commands.
- styles/Kyberforge/VagueQualifier.yml: prune 17 tokens verified
false-positive-dominated on this repo's own voice via a real corpus
sweep (obvious, clearly, usually, several, simple, easy, completely,
simply, tiny, etc.), keep 13 with real or unattested noise. Revert the
28 prose "fixes" those tokens drove across 14 skill files back to their
original, correct wording, including a functional regression to
caveman/SKILL.md's own filler-word list (a mention, not a use) — now
guarded with vale-off comments against recurrence.
Gaps:
- --minAlertLevel=warning on the pre-commit hook and Step 1 invocation
so warning-level rules actually surface, without collapsing the
FAIL/SUGGESTION severity mapping skill-audit/agent-audit rely on.
- vale-wrap.sh: fix --config=<path> equals-form, absolute-path silent
no-op, and a zero-file-argument stdin hang.
- Route vale-run and lint-runner through a documented wrapper script
when a target repo has one, instead of unconditionally recommending
bare `vale`.
- Wire Kyberforge.VagueQualifier/SentenceOpenerThereIs into skill-audit/
agent-audit's dimension-mapping prose (Body discipline).
- Add plugins/lint/sources.md provenance for lint-runner (ADR-0010).
- Sync both marketplace.json lint-entry descriptions with plugin.json.
- Retune skill-size-check.sh's MAX_WORDS 5000->2900 (measured ~1.6-1.7
tokens/word on this repo's corpus, the old value gated at ~8,500
tokens against a stated 5,000 ceiling); fix the >/>= line-count
boundary and wc -l undercount on files with no trailing newline.
- Document the vale binary as a Setup prerequisite in AGENTS.md.
- Fix SentenceOpenerThereIs's dead regex alternative and add a real
sentence-start anchor/scope.
- Fix a stale docs/research/docs/vale/ index pointer in kyberforge's
docs README (moved to plugins/lint/ in e1a5403).
- Rewrite ADR-0013's Consequences section past-tense to describe what
actually landed, and record the styles-portability limitation
(repo-root placement stays intentional; deferred to a separate
session per this PR's review).
Test coverage: 9 new vale-wrap.sh fixtures (quotes, backslash/unicode,
blank-line paragraphs, --config= form, zero-arg/absolute-path handling,
literal-block no-regression) and boundary-pair tests for
skill-size-check.sh's line/word ceilings.
bash tests/run-tests.sh: 9 scripts + 125 bats assertions, all passing.
scripts/check-manifests.sh and claude plugin validate --strict: clean.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MCQ648fLSFXPHGZdQ8gn58
Deferred item from PR #85 review. Per ADR-0013: cherry-picks two low-noise
rules from trialing write-good/alex against the real corpus (VagueQualifier,
SentenceOpenerThereIs) into styles/Kyberforge rather than adopting either
package wholesale (both are tuned for blog prose and were noisy on this
repo's terse, imperative instruction files - see the ADR's rejected-rule
list). Adds a new skill-size-check pre-commit hook enforcing agentskills.io's
500-line/5,000-token SKILL.md ceiling, currently unenforced. Fixes the 28
resulting violations across 20 existing SKILL.md/agent files so the
enforcing pre-commit hook lands clean.
governance.md/CONTROLS.md were evaluated and excluded as rule sources -
they're org/CI-infrastructure controls, not prose patterns Vale can express.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01QUDczvw1H3eEeMD29Q9Lbi
Vale's text.frontmatter.description scope silently stops matching once
the description is a YAML block scalar spanning 2+ physical lines —
the style used by most skills/agents in this repo. scripts/vale-wrap.sh
flattens the description to one line in a scratch copy (preserving the
repo-relative path and total line count) before invoking real vale, and
both audit skills plus the pre-commit hook now call it instead of vale
directly. Also tightens the pre-commit hook's file glob to single path
segments so it can't cross into docs/research examples or asset
templates the way the audit skills' scoped invocations already avoid.
Addresses PR #85 review feedback.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FxG5T8EJDgkABXxuneuFfn
Adds repo-root .vale.ini plus a custom Kyberforge style (description-opener,
vague-wording, and generic reference-pointer padding rules) and a
KyberforgeCopilot style scoped to .agent.md files (Use proactively check).
skill-audit and agent-audit Step 1 now run vale against the specific file(s)
being audited and defer the corresponding Description/Patterns/Body checks
to its output instead of re-deriving them by LLM judgment, per the split
proposed in issue #84.
Closes#84
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FxG5T8EJDgkABXxuneuFfn
