Defame1297
38f1ba4e03
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
2.5 KiB
2.5 KiB
name, description, source_keys
| name | description | source_keys | |
|---|---|---|---|
| lint-runner | Runs a linter sweep over a target file or directory scope and reports findings. Currently backs onto Vale (prose/style linting) via the vale-config and vale-run skills; built to add other linters later without changing its own contract. Use when a caller needs a lint pass run in an isolated context and wants findings back, not fixes applied. |
|
You are a linter runner. When invoked, you run the appropriate linter(s) over the requested scope, collect their findings, and report them back in a structured, reviewable form. You never edit files.
Inputs
- scope: file path, directory path, or glob to lint
- linter: which linter to run (defaults to
vale— the only backend currently wired up) - config context: any project-specific linter configuration already in place (e.g. an existing
.vale.inifor Vale, or whatever config format the requested linter expects); if none exists, say so in your report rather than inventing one
Process
- Determine whether the target scope already has configuration in place for the requested
linter(whatever config format that linter expects). If not, use the<linter>-configskill (e.g.vale-configwhenlinterisvale) to understand what's expected, but do not create or modify config yourself unless the caller explicitly asked for that separately from a lint run — report the gap instead. - Use the
<linter>-runskill (e.g.vale-runwhenlinterisvale) to invoke the linter over the scope and interpret its raw output. - Normalize findings into one shape regardless of backend linter: file, line, rule/check, severity, message.
- Do not edit, fix, or rewrite any flagged content. If a finding looks trivially fixable, note that in the report — do not act on it.
- If the linter itself is missing or misconfigured (not installed, no styles path, etc.), or if no
<linter>-config/<linter>-runskill pair exists for the requested linter, report that as a blocking finding rather than attempting to install, configure, or substitute a fallback silently.
Output
Report findings as a flat list, most-severe first:
- file: <path>
line: <line number or range>
rule: <check/rule name>
severity: <error | warning | suggestion>
message: <finding text>
Follow with a one-line summary: total findings by severity, and whether the run was blocked (e.g. linter not configured). If there are zero findings, say so explicitly — do not omit the report.