Agent Skills

A personal collection of Agent Skills — portable, self-contained capabilities that any compatible AI agent can discover and load on demand. Built on the open standard originally published by Anthropic, now adopted across Claude, Codex CLI, Gemini CLI, GitHub Copilot, Cursor, and 25+ other platforms.

Each skill is a folder containing a SKILL.md file (instructions + metadata) and, where needed, supporting scripts/, references/, or assets/. Nothing here is Claude-specific unless explicitly noted — see CONTRIBUTING.md for the portability rules this repo follows.

Important

Stalled-work tracking. Tasks blocked on external human action (a DUA signature, a sign-up, a records request) are tracked as GitHub issues and surfaced weekly — see docs/stalled-work-tracking.md for how to open, resolve, and check on one.

Jump to: Skills · Installing · Repo structure · Wiki system

Skills

Skill	Category	Description
agent-eval	Agent Design	Designs and runs evaluations for LLM/agent outputs — rubrics, LLM-as-judge scoring, regression test sets, and pass-rate reporting with a runnable scoring script.
agent-redteam	Agent Design	Generates adversarial test cases for safe-failure testing — refusals, hedging, graceful degradation. Pairs with agent-eval for scoring.
deid-reid-harness	Agent Design	Adversarial de-identification ⟷ re-identification eval harness for clinical text — generates synthetic notes with ground-truth PHI spans, runs a de-id pipeline, and scores Safe Harbor leakage, Expert Determination re-id risk, and free-text inference across a privacy-utility frontier. Model-independent, verified offline with bootstrap CIs and significance tests. (sample results)
repo-pincer	Agent Design	Reverse-engineers a codebase by reconciling top-down claims (docs, README, API surface) against bottom-up reality (actual implementation) — classifies every claim as Confirmed, Drift, Aspirational, or Silent. Runs standalone; optionally compiles into a wiki vault.
wiki-operator	Knowledge Management	On-demand vault operations — /learn, /update, /connect, /ask, /review, /quiz, /map, /source, /clean, /health. The primary interface for working with the wiki. Requires Obsidian MCP connected.
wiki-synthesizer	Knowledge Management	Batch compilation — automatically preprocesses unstructured journals, promotes ideas into concept pages, compiles Sources/raw/ into source pages, updates the hot cache. Run after learning sessions. Requires Obsidian MCP connected.
wiki-librarian	Knowledge Management	Structural maintenance — audits broken links, orphans, stale notes, duplicates, and contradictions. Proposes fixes with confirmation. Run weekly. Requires Obsidian MCP connected.
wiki-governor	Knowledge Management	Self-governing maintenance loop — orchestrates the librarian and synthesizer, then adds a constitution-compliance audit, a tracked health score, and a knowledge-gap queue. Keeps the vault accountable to its own rules. Requires Obsidian MCP connected.
wiki-warehouse	Knowledge Management	Cold storage for raw documents — ingests PDFs/ebooks/scans into a separate private GitHub repo (intake.py: hash → extract text, OCR fallback for scans → manifest), then writes a lean content-hash pointer note into the vault. Keeps originals and full text out of the vault. /ingest, /warehouse-audit. Requires Obsidian MCP + the warehouse repo cloned.

Installing a skill

Claude Code (plugin marketplace):

/plugin marketplace add Hefrock/agent-skills
/plugin install agent-eval@hefrock-agent-skills

Install any other skill the same way — swap agent-eval for the plugin name from the table above (e.g. /plugin install wiki-governor@hefrock-agent-skills).

Updating after new skills are added to this repo:

/plugin marketplace update hefrock-agent-skills

Run this before installing anything added since your last update — it refreshes the marketplace's list of available plugins. Already-installed plugins pick up content changes (new commands, edited SKILL.md) automatically; a plugin that's new to the marketplace still needs its own /plugin install <name>@hefrock-agent-skills afterward. If a fresh plugin still comes back "not found" right after marketplace update, restart the CLI session and retry.

Claude Code (manual, no plugin system):

git pull origin main   # if you already have the repo cloned
cp -r skills/agent-eval ~/.claude/skills/

claude.ai: Zip the individual skill folder (e.g. skills/agent-eval/) and upload via Settings → Features → Custom Skills (requires a paid plan with code execution enabled).

Other platforms (Codex, Gemini CLI, Cursor, etc.): Copy the skill folder into whatever directory that platform scans for skills — the SKILL.md format works unmodified.

Repo structure

Expand directory tree

agent-skills/
├── .github/
│   ├── workflows/ci.yml        # builds + tests the MCP server, runs the Python skill suites
│   └── ISSUE_TEMPLATE/         # blocked-human / dated-followup templates for stalled-work tracking
├── .claude-plugin/
│   └── marketplace.json        # Claude Code-only install metadata — optional, additive
├── docs/
│   └── stalled-work-tracking.md # how the blocked-human / dated-followup convention works
├── skills/                     # flat — one folder per skill, no category nesting
│   ├── agent-eval/             # rubric-based evals, LLM-as-judge, regression test sets
│   ├── agent-redteam/          # adversarial case generation, pairs with agent-eval
│   ├── deid-reid-harness/      # clinical de-id/re-id eval — scripts, refs, 31-test suite
│   ├── repo-pincer/            # codebase reverse-engineering — claims vs. reality reconciliation
│   ├── wiki-operator/          # on-demand vault operations
│   ├── wiki-synthesizer/       # journal preprocessing + concept page compilation
│   ├── wiki-librarian/         # structural health audits — scripts/check_vault.py, 23-test regression suite
│   ├── wiki-governor/          # maintenance loop + constitution compliance + health score
│   └── wiki-warehouse/         # raw-document cold storage (external repo) + vault pointers
├── mcp/
│   └── obsidian-vault/         # MCP server required by wiki-operator
│       ├── src/index.ts        # 10 tools: search, read, write, append, patch, query, links, delete
│       ├── test/               # end-to-end STDIO tests — `npm test`
│       └── README.md           # setup and configuration guide
├── knowledge-os/
│   ├── constitution.md         # 10 laws Claude follows when operating the wiki
│   ├── architecture.md         # component map, data flow, note lifecycle
│   └── sitrep.md               # living status + gap analysis for the wiki system
├── templates/                  # note templates copied into vault by setup-vault.sh
│   ├── concept.md
│   ├── journal.md
│   ├── source.md
│   └── map.md
├── bin/
│   └── setup-vault.sh          # one-command vault bootstrap (creates folders, copies templates + constitution)
├── skill-template/              # starting point for a new skill
└── CONTRIBUTING.md             # how to add a skill, including portability rules

Wiki system

The wiki skills (wiki-operator, wiki-synthesizer, wiki-librarian, wiki-governor, wiki-warehouse) form a complete personal knowledge system built around an Obsidian vault. wiki-warehouse adds a separate private "cold storage" repo for raw documents, keeping originals out of the vault while indexing them by content-hash pointer.

Every wiki skill requires the obsidian-vault MCP server connected — nothing works without it. To enable it:

1. Build it once:

   cd mcp/obsidian-vault && npm install && npm run build

2. Point it at your vault. Three ways, easiest first:
   • Claude Code CLI (recommended):

     claude mcp add obsidian-vault \
       -s user \
       -e OBSIDIAN_VAULT_PATH=/absolute/path/to/your/vault \
       -- node /absolute/path/to/agent-skills/mcp/obsidian-vault/dist/index.js

     -s user registers it at the user level — available in every project, which is how this server is meant to run — and edits ~/.claude.json for you. This avoids hand-editing that file directly, which can get weird in a GUI editor if a running Claude Code process has it open.
   • setup-vault.sh — run ./bin/setup-vault.sh ~/path/to/vault; it bootstraps the folder structure, copies templates and the constitution into System/, and prints a config snippet with your paths filled in, for the manual route below.
   • Manual — add this to ~/.claude.json yourself:

     {
       "mcpServers": {
         "obsidian-vault": {
           "command": "node",
           "args": ["/absolute/path/to/agent-skills/mcp/obsidian-vault/dist/index.js"],
           "env": { "OBSIDIAN_VAULT_PATH": "/absolute/path/to/your/vault" }
         }
       }
     }

3. Restart Claude Code and verify: run /mcp — expect obsidian-vault connected with 10 tools. Full tool reference: mcp/obsidian-vault/README.md.

This is a per-device setup step — the server is a local process, so a new machine needs its own build and its own ~/.claude.json entry (with paths for that machine), even if it's pointed at the same synced vault.

The operating layer lives in this repo:

Path	Purpose
knowledge-os/constitution.md	10 non-negotiable rules Claude follows when operating the wiki
knowledge-os/architecture.md	Component map, data flow, and note lifecycle reference
knowledge-os/sitrep.md	Living status + gap analysis for the wiki system — what's shipped, what's untested, what's next
templates/	Note templates (concept, journal, source, map) — copied into System/templates/ in your vault by setup-vault.sh

License

See LICENSE.