cf68841220
* feat(#1243): consume Claude plugin-provided skills via native Skill-tool directive + grant Skill to agent_skills-consumer agents - Relax global skill name validation to accept namespaced form `^[A-Za-z0-9_-]+(:[A-Za-z0-9_-]+)*$` - Namespaced names (containing colon) on claude runtime emit a Skill-tool load directive instead of a @-include line - Namespaced names on non-claude runtimes are skipped with a warning - Bare unresolved names retain existing warn-and-skip behavior (no promotion to directive) - Grant `Skill` tool to all 22 agent_skills consumer agents; 5 generated agents updated via research-profiles.cjs + regen, 17 hand-authored agents edited directly - Add 16 TDD tests in describe('bug #1243') covering happy/mixed/precedence/negative/cross-runtime/regression/grant cases Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * docs(#1243): document plugin-provided skills in agent_skills Update the Agent Skills Injection reference in CONFIGURATION.md with the three entry forms (project-relative, global:<name>, global:<plugin>:<skill>), the Claude-only runtime behaviour of the namespaced form and the warn-skip on other runtimes, the plugin pre-install prerequisite, and the consumer-agent Skill tool grant. Add docs/how-to/attach-a-plugin-skill-to-a-gsd-agent.md with a step-by-step guide for installing the plugin, locating the namespaced skill name, wiring it into agent_skills, and verifying injection. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * fix(#1243): align agent_skills docs with emitted block format + mixed-block regression test (code-review) - Replace two-section mixed-block example (bogus "Load these plugin-provided skills using the Skill tool:" header) with the actual single-section inline format in CONFIGURATION.md and docs/how-to/attach-a-plugin-skill-to-a-gsd-agent.md - Fix quoted warning text in how-to doc to exactly match the emitted string: [agent-skills] WARNING: Plugin-namespaced skill "global:<name>" requires a Skill-tool-capable runtime (claude) — skipping on runtime "<runtime>" - Replace phantom agent slugs (gsd-checker, gsd-researcher, gsd-advisor, gsd-synthesizer) in CONFIGURATION.md Supported Agent Types with real agents/gsd-*.md examples (gsd-plan-checker, gsd-phase-researcher, gsd-code-reviewer, gsd-ui-auditor, gsd-research-synthesizer) - Add byte-identical mixed-block regression test: one path-resolvable global skill + one plugin-namespaced skill on claude runtime → asserts r.ir.block === single-section interleaved block, no secondary header Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * chore(#1243): regenerate agent-size baseline for the Skill-tool grant The 22 agent_skills-consumer agents each grew +7 bytes from adding `Skill` to their tools list; refresh the committed per-agent size baseline (#1074 guard). * chore(#1243): add Added changeset fragment * fix(#1243): traceable allow-test-rule ref + separator-agnostic byte-identical tests (CI) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> --------- Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
99 lines
4.0 KiB
Markdown
99 lines
4.0 KiB
Markdown
# How to attach a plugin-provided skill to a GSD agent
|
|
|
|
Use this guide when you have a Claude Code plugin that ships a skill and you want GSD agents to load it automatically at spawn time. The `global:<plugin>:<skill>` entry form makes GSD emit a Skill-tool load directive so the agent picks up the plugin skill alongside any project-local or personal global skills.
|
|
|
|
**This feature works on the Claude runtime only.** On other runtimes (Codex, Gemini, Cursor, etc.) the entry is skipped with a warning. Plan accordingly if your team uses multiple runtimes.
|
|
|
|
---
|
|
|
|
## 1. Install the plugin in Claude Code
|
|
|
|
GSD does not install or manage Claude Code plugins. Install the plugin first:
|
|
|
|
```
|
|
/plugin install <plugin-source>
|
|
```
|
|
|
|
Verify the plugin is active and note the skill names it exposes. The plugin's documentation or `README` will list available skill names in the format `plugin:skill` — for example, `coderabbit:code-review`.
|
|
|
|
---
|
|
|
|
## 2. Find the namespaced skill name
|
|
|
|
The name you need has two colon-separated segments after `global:`:
|
|
|
|
```
|
|
global:<plugin>:<skill>
|
|
```
|
|
|
|
For example, if the plugin slug is `coderabbit` and it provides a skill called `code-review`, the entry is:
|
|
|
|
```
|
|
global:coderabbit:code-review
|
|
```
|
|
|
|
Each segment must consist of alphanumeric characters, underscores, or hyphens only. No spaces, no slashes, no dots.
|
|
|
|
---
|
|
|
|
## 3. Add the entry to `agent_skills`
|
|
|
|
Open `.planning/config.json` and add the namespaced skill to the array for the agent type that should receive it:
|
|
|
|
```json
|
|
{
|
|
"agent_skills": {
|
|
"gsd-executor": [
|
|
"skills/project-conventions",
|
|
"global:coderabbit:code-review"
|
|
],
|
|
"gsd-verifier": [
|
|
"global:coderabbit:code-review"
|
|
]
|
|
}
|
|
}
|
|
```
|
|
|
|
You can mix all three entry forms freely within the same array — project-relative paths, `global:<name>` personal skills, and `global:<plugin>:<skill>` plugin skills.
|
|
|
|
Or set it from the CLI:
|
|
|
|
```bash
|
|
gsd-tools query config-set agent_skills.gsd-executor '["skills/project-conventions","global:coderabbit:code-review"]'
|
|
```
|
|
|
|
---
|
|
|
|
## 4. Verify the injection
|
|
|
|
Start a phase on the Claude runtime. When the target agent is spawned, its Task() prompt will contain an `<agent_skills>` block that includes a Skill-tool load directive for the plugin skill:
|
|
|
|
```xml
|
|
<agent_skills>
|
|
Read these user-configured skills:
|
|
- @skills/project-conventions/SKILL.md
|
|
- Load the `coderabbit:code-review` skill via the Skill tool before proceeding (plugin-provided).
|
|
</agent_skills>
|
|
```
|
|
|
|
Both entry types appear in the same `Read these user-configured skills:` section, interleaved in config order. There is no separate header for plugin-provided skills.
|
|
|
|
If you see `[agent-skills] WARNING: Plugin-namespaced skill "global:coderabbit:code-review" requires a Skill-tool-capable runtime (claude) — skipping on runtime "<runtime>"` in the logs, the entry is being skipped because the active runtime is not Claude. The configuration is still valid; no change is needed.
|
|
|
|
---
|
|
|
|
## Notes
|
|
|
|
- **Plugin must be pre-installed.** GSD only references the skill by namespaced name. If the plugin is not installed in Claude Code, the Skill tool call will fail at agent runtime — GSD cannot validate plugin presence at configuration time.
|
|
- **Which agents support it.** All 22 GSD agent types that consume `agent_skills` carry the `Skill` tool, so any of them can load plugin-provided skills. See [Agent Skills Injection — Supported Agent Types](../CONFIGURATION.md#supported-agent-types) for the full list.
|
|
- **Non-Claude runtimes.** The `global:<plugin>:<skill>` entry is silently skipped with a warning on non-Claude runtimes. Project-relative and `global:<name>` entries work on all runtimes.
|
|
- **Security boundary.** Plugin skill content is resolved entirely by Claude Code at agent runtime. GSD does not read, cache, or validate the plugin's files.
|
|
|
|
---
|
|
|
|
## Related
|
|
|
|
- [Configuration — Agent Skills Injection](../CONFIGURATION.md#agent-skills-injection)
|
|
- [Install a minimal GSD and add skills later](install-minimal-and-add-skills.md)
|
|
- [Import a capability from a URL](import-a-capability-from-a-url.md)
|