cmc-27/pass-the-parcel/wiki-query
Claude Code and agent skill by cmc-27/pass-the-parcel
- community
Checking that the SKILL.md file is still in the repository.
Use when asking a question about the codebase, looking up a concept, or synthesizing an answer from the wiki + ref/ docs. Triggers: 'what do I know about X', 'look up Y', 'find in wiki', 'search docs', 'wiki says'. Read-only — defers writing to knowledge-capture.
Install
npx vskill@latest install cmc-27/pass-the-parcel/wiki-querySKILL.md
Wiki Query
Goal
Answer questions by synthesising the contents of .wiki/ (including ref/). Cite every claim with [Title](path) so the user can verify. Never write to disk. If the answer should be persisted, point the user to the knowledge-capture skill.
Workflow
1. Build the Index Map
- Read
.wiki/core/00-system-index.mdto enumerate all wiki docs with summaries. - Read
.wiki/ref/ref-index.mdto enumerate all reference docs with topics. - Together these form the candidate pool for the question.
2. Match Question to Candidates
- Identify keywords in the question (e.g. "banker's rounding", "NZS 3910", "price breakdown").
- Grep the wiki for those keywords to find candidate docs.
- For topic-keyword matches, prefer
ref/docs (e.g. tag writing →REF-Tag-Schedule.md).
3. Read & Synthesise
- Read the top 3-7 most relevant candidates (1-2 from
ref/if topic-matched, rest from main wiki). - If after reading the top 7 candidates NO doc covers the question, output exactly:
> "No relevant docs found in the wiki. The question may be out-of-scope, or you may want to add knowledge via the knowledge-capture skill first." Do NOT fabricate an answer from training data when no source is found.
- Otherwise synthesise a chat answer with
[Title](relative-path)inline citations.
4. Cite Format
In-conversation citations use project-root-relative paths:
[Article Title](.wiki/core/18-knowledge-capture.md)[Tag Schedule](.wiki/ref/REF-Tag-Schedule.md)
Within a citation list at the bottom of the answer, prefer the same project-root-relative form.
5. End With Persistence Pointer
Always close the response with:
"Want to persist this insight as a knowledge-capture decision? Use the@knowledge-captureskill to add an entry."
This is a pointer, not an action. The user decides whether to write.
6. Out-of-Scope Behaviour
- Code questions (e.g. "what does
handleClickdo?") — out of scope. Point togrep/ codebase exploration, not the wiki. - Recent changes (e.g. "what changed in the last deploy?") — out of scope. Point to
.devops/logs/agent-changelog.md. - Knowledge operations (e.g. "add this to the wiki") — out of scope. Point to
@knowledge-captureor@wiki-lintskills.
Usage Guidelines
- Read-only by design. This skill never writes to disk. If you find yourself wanting to save the answer, that's a separate skill (
@knowledge-capture). - Prefer the wiki over training data. When a doc covers the question, cite it. Do not answer from memory when a source is available.
- Be honest about gaps. "No relevant docs found" is a valid, useful answer.
Security
- Security scanPassed, 95 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryChecking.devops/skills/wiki-query/SKILL.md
Popularity
- 7-day trend score
- 5
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/cmc-27/pass-the-parcel/wiki-query)