holocron/plugins/kyberforge/skills/skill-author/references/retrofit.md

source_keys

source_keys
agentskills-best-practices agentskills-optimizing-descriptions

Retrofitting a skill to the ADR-0020 contract

Read this when references/improve.md Step 4 sends you here: the skill you are editing is over the description or body budget and has to come into contract before any other change can be committed. The gates are hot and carry no baseline file, so a one-line fix to a non-compliant skill is blocked until this is done.

Measure first. Do not guess which gate fired: run /skill-audit on the directory and read its ### Structure dimension, which reports the description characters and the body-only word count separately from the whole-file spec backstop. Retrofit against the number that actually fired — a skill can sit a thousand words inside the whole-file backstop while failing the body budget.

Validate in place. Audit the skill's real directory inside its package. Never audit a copy in a scratch directory, and never move a skill out to work on it: the boundary-target universe is built by walking up from the file being checked, so a copy with no authoring root above it resolves against nothing and the check declines rather than running —

INFO  boundary-target resolution DID NOT RUN — no skill universe could be determined for
this path ... Unchecked target(s): totally-fake-target

The run still exits 0, so that line reads as a pass and is not one. Treat DID NOT RUN as not checked, always. A retrofit signed off on a scratch copy carries an unverified boundary target into the corpus, which is precisely the failure this gate exists to catch.

Cut in this order

Work the list top down and stop as soon as the gate clears. The order is by ratio of tokens removed to behaviour lost — inverting it is how a retrofit ends up deleting the one instruction the skill existed to carry.

1. Gotchas that paraphrase a step in the body below. Zero information, and already a FAIL on its own. Delete the Gotcha, keep the step.
2. Spec restatements — text that repeats a published specification, a tool's --help, or a ceiling the validator already enforces. The agent gets this right without it. Delete, or move the table to references/ if a flow genuinely needs to look it up.
3. Capability enumeration — in a description, the feature list after the trigger clause; in a body, the paragraph that recites what the skill can do. One capability clause survives in the description; the rest belongs in README.md.
4. Per-flow prose — anything only one branch of the procedure ever reaches. This is the largest single win in most bodies, and it is a move, not a delete: each flow gets its own self-contained references/ file, wired from a dispatch table.

If the body is still over after all four, the skill is doing two jobs. Split it, and say so rather than compressing prose until it stops being readable.

What "mutually exclusive flows" means

Two or more flows that a single invocation cannot both take. The three-way test, copied verbatim from the body-discipline rubric /skill-audit judges against — nothing to load, it is quoted in full here:

separate subcommands, separate input types, separate lifecycle stages

Any one of the three is enough. Two flows that differ only in a parameter value are one flow. At two or more mutually exclusive flows a dispatch table is mandatory regardless of word count, because every invocation otherwise pays for every branch it did not take.

Reference-file conventions

The create flow owns these rules, and this flow is forbidden from reading references/create.md, so what a retrofit needs is restated here:

• One topic per file. A file mixing two concerns gets loaded for one of them and spends the caller's context on the other.
• Kebab-case filenames, named after the topic rather than the flow that reads it — body-discipline.md, not step-3.md.
• Wire every file with the literal conditional form If <condition>, read `references/<file>.md` . A generic pointer ("see references/ for details") is a Vale error.
• Two hops from SKILL.md, never three. A flow file may route on to a shared contract file; a file reachable only through two intermediates is rarely loaded when it is needed.
• source_keys frontmatter. If the content you are moving drew on a research source, the new file needs top-level source_keys: frontmatter listing those slugs, and every slug must already exist as an ## <slug> heading in references/sources.md. Moving sourced content out of SKILL.md without carrying its slugs across breaks the provenance chain, and /skill-audit reports the new file as an INFO with no source_keys.

Collateral is mandatory, not optional

Moving content out of a SKILL.md leaves three files describing a structure that no longer exists. /skill-audit's provenance check exits clean on all three of these, so nothing catches them for you. After every retrofit that adds, removes or renames a file:

• README.md file table — a row for every new references/ file, and no row left for a file that is gone. Say what triggers the load, not just what the file contains.
• references/README.md, where the skill has one — same update, same reason.
• references/sources.md → Contributing files — add the new file to every slug whose content moved into it, and remove any file the retrofit deleted. This is the one that gets missed: sources.md keeps citing sections of SKILL.md that no longer exist, the provenance check still exits 0, and the stale claim survives review.
• Reachability of every relocated gate. For each Gotcha or gate the retrofit moved out of the body, list the flows that need it and confirm each one reaches the surviving copy. A gate that lands in a single flow file is invisible to every other branch, and no gate detects that: /skill-audit reads whichever file it was handed, and the word counts improve either way. Where more than one flow needs it, the copy belongs in the body's common-gates section, not in a flow file. Grep the skill for the gate's key term and check every branch that hits zero.
• Re-run /skill-audit and confirm its ### Provenance dimension does not report the new file as missing source_keys.

Compression must not add authority the source text didn't have

This one is not part of the checklist above, and deliberately so: it fires on a wording change with no file change at all, so a retrofit that adds and removes nothing still owes it.

The sources.md bullet above is about an entry going stale — Contributing files left uncited after content moves. This is a distinct failure: a compression or rewrite pass that upgrades an honest hedge in a Description into an unsupported confident claim, without the underlying source having changed at all — "no forge-specific content drawn directly from it beyond that" quietly becoming "Grounds Step 2's dispatch table."

/skill-audit's provenance script does now notice this class: it diffs each slug's Description and Contributing files text against a base ref and raises an INFO when the wording changed. That is a prompt, not a verdict — it reports only that the claim moved, never whether the new claim is true, because a bash script can verify an entry is internally consistent and nothing more. Answering it is this flow's job: if a retrofit strengthens or otherwise changes the wording of a provenance claim, re-read the upstream research doc first and confirm the stronger wording is actually still true before committing it.

Versioning a retrofitted skill

SKILL.md Step 4 says to bump the patch version on improve, which presumes there is a version to bump. A pre-ADR-0020 skill often carries none — metadata.version only became mandatory under ADR-0022, and this flow is exactly where those skills surface.

A skill with no metadata.version is seeded at "1.0.0", not bumped. "0.1.0" is reserved for a skill created new by the create flow: it means "created and never yet revised", which understates a skill that has been through retrofit and audit passes without tracking a version. Add the field in this retrofit — the skill-frontmatter pre-commit hook blocks the commit without it.

Worked example — a description retrofit

gitea-issues before, 827 characters, the single most common shape in the corpus:

Use when reading or writing Gitea issues: listing repo issues, getting a single issue's details/
comments/labels, creating an issue, updating its state, adding or editing comments, applying
labels via issue_write, or searching issues/PRs across repositories. Triggers on "create an
issue", "what issues are open", "get issue #N", "close issue #N", "comment on issue #N", "search
issues for X" — even when the user doesn't say "Gitea" explicitly. Composes gitea-labels-
milestones for all label inference/resolution and milestone lookup — do not use this skill to
manage label or milestone definitions themselves (create/edit/delete a label, create/close a
milestone), that's gitea-labels-milestones directly. Do not use for pull requests (use gitea-prs)
or for local git branch/commit work (use gitea-branches or git-branches).

After, the 290 characters that shipped:

Use when reading or writing Gitea issues — "create an issue", "what issues are open", "close
issue #N", "comment on issue #N", "search issues for X" — even when the user does not say
"Gitea". Not pull requests -> `gitea-prs`. Not label or milestone definitions ->
`gitea-labels-milestones`.

The retrofit kept the quoted-phrasing register and dropped the verb list, not the other way round. Either register is admissible — what is banned is carrying both. Choose whichever routes better for the skill in hand; here the quoted user phrasings do, because they are how people actually ask.

What came out, and why:

Removed	Why
The second trigger register — Triggers on "create an issue", "what issues are open", …	The same triggers restated as quoted user phrasings. Two registers of one trigger list is a FAIL, not a suggestion.
applying labels via issue_write	Implementation detail. The router does not choose a skill by which MCP call it makes.
Composes gitea-labels-milestones for all label inference/resolution and milestone lookup	A composition note. It changes no routing decision and belongs in README.md.
The parenthetical (create/edit/delete a label, create/close a milestone)	Capability enumeration inside a boundary clause. The boundary needs the target, not its feature list.
The gitea-branches / git-branches boundary	Dropped entirely. Neither was ever going to win an issue request, so the clause defended against nothing — an invented boundary costs characters and buys no routing accuracy.
Do not use for pull requests (use gitea-prs) prose form	Kept, but rewritten as Not pull requests -> \gitea-prs`.` The rewrite buys characters, one uniform shape for the router, and a stricter check: an unresolved arrow target is a blocking ERROR, while an unresolved prose target is only a SUGGESTION unless another target in the same sentence resolves. The prose form does not dangle as loudly.

What stayed: one trigger clause, one capability clause, the indirect trigger (genuinely warranted here — people say "create an issue", not "create a Gitea issue"), and the boundary clauses.

Two rules the gates enforce but the prose does not spell out

Boundary clauses may be plural. Write one per genuine near-miss — the example above carries two, because two different skills could each steal activations. "A boundary clause" in the contract means at least one, not exactly one. What is banned is a boundary clause invented for a skill that was never going to compete, not a second real one.

Never let a hyphenated routing target wrap across lines in a folded > scalar. YAML folding replaces the newline with a space, so gitea-labels- at the end of one line and milestones at the start of the next fold into gitea-labels- milestones. The gate then reads the target as gitea-labels, finds no such skill, and reports a dangling boundary target. This is not hypothetical — it is how gitea-labels-milestones broke (issue #100). It is fixed: the corpus carries no dangling target today, and the repo's test suite pins that set as empty, so a reintroduction fails the suite rather than joining a backlog. Reflow the line so the whole name sits on one of them. The same applies to any backticked skill or agent name in a description.