AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified MIT Self-run

Research

skill-ivankuznetsov-llm-wiki-research · by ivankuznetsov

Search QMD-indexed project and main cross-project LLM wikis for past patterns, decisions, solutions, and pitfalls before planning or implementation. Use before feature planning, bug fixing, refactoring, architecture work, or code changes when a project wiki may exist.

No reviews yet
0 installs
0 views
view→install

Install

$ agentstack add skill-ivankuznetsov-llm-wiki-research

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-ivankuznetsov-llm-wiki-research)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
2d ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Research? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Wiki Research

Search the project wiki and main cross-project wiki before planning or implementation. Produce a concise Past Knowledge section that downstream planning or coding can use.

Step 1: Identify Context

Determine:

  • Current project name: basename "$(git rev-parse --show-toplevel 2>/dev/null || pwd)"
  • Parent of project: dirname "$(git rev-parse --show-toplevel)"
  • Whether wiki/ exists in the repo.
  • Whether .llm-wiki/config.json exists and contains main_wiki_path.
  • Whether ~/wikis/master/wiki/ exists.
  • Whether ~/wikis/main/wiki/ exists.
  • Whether /wikis/master/wiki/ exists.
  • Whether /wikis/main/wiki/ exists.
  • Whether QMD MCP tools, qmd CLI, or only rg are available.

If no main cross-project wiki exists, say that explicitly. Ask the user for a main wiki folder when cross-project context is important, or suggest running bootstrap to create /wikis/master/wiki/.

Step 2: Search With the Best Available Tool

Preferred: QMD MCP

Use at most two QMD MCP query calls. Do not fire many QMD calls in parallel; combine related searches into the searches array.

Project search:

{
  "searches": [
    { "type": "lex", "query": "" },
    { "type": "vec", "query": "" }
  ],
  "collections": [""],
  "limit": 10
}

Main cross-project search:

{
  "searches": [
    { "type": "lex", "query": "" },
    { "type": "vec", "query": "patterns for " }
  ],
  "collections": ["master"],
  "limit": 10
}

Use the master QMD collection for the main cross-project wiki when available. If the environment uses a main collection instead, query main.

Read the top 3-5 relevant pages with QMD get or multi-get.

Fallback: QMD CLI

qmd query "" --collection ""
qmd query "" --collection master
# or, if the cross-project wiki is indexed as main:
qmd query "" --collection main

Fallback: ripgrep

rg "" wiki/ --type md -C 2
rg "" "" --type md -C 2
rg "" ~/wikis/master/wiki/ --type md -C 2
rg "" ~/wikis/main/wiki/ --type md -C 2
rg "" ../wikis/master/wiki/ --type md -C 2
rg "" ../wikis/main/wiki/ --type md -C 2

Skip missing directories gracefully.

Step 3: Read Important Pages

Prioritize:

  • wiki/decisions.md
  • wiki/technical-debt.md
  • wiki/gaps.md
  • wiki/architecture.md
  • /patterns.md
  • /learnings.md
  • ~/wikis/master/wiki/patterns.md
  • ~/wikis/master/wiki/learnings.md
  • ~/wikis/main/wiki/patterns.md
  • ~/wikis/main/wiki/learnings.md
  • /wikis/master/wiki/patterns.md
  • /wikis/master/wiki/learnings.md
  • /wikis/main/wiki/patterns.md
  • /wikis/main/wiki/learnings.md

Only cite or summarize pages actually read.

Step 4: Produce Past Knowledge

Return this structure:

### Past Knowledge

**Relevant wiki pages:**
- ...

**Applicable patterns:**
- ...

**Past decisions that constrain this work:**
- ...

**Known pitfalls:**
- ...

**Reusable components:**
- ...

**Gaps this work could fill:**
- ...

If no wiki exists, say that explicitly and suggest running bootstrap.

Rules

  • Search wiki context before codebase exploration when wiki context exists.
  • Do not claim wiki facts unless a page was read.
  • Prefer concise synthesis over dumping search results.
  • Preserve uncertainty and missing coverage as gaps.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.