lich-skills

lich-skills

Personal skill collection for Claude Code, Gemini CLI, and OpenAI Codex — by @LichAmnesia.

Telegraph-style, opinionated, no filler. Engineering judgment skills plus high-leverage utility tools.

中文版：README-zh.md

---

The go-no-go gate (Stage 0)

                          ┌─────────────────────────────────────┐
                          │   NO-GO is the default.             │
                          │   GO requires evidence.             │
                          └──────────────┬──────────────────────┘
                                         │
   MEMORY     →   HYPOTHESIS   →   FIVE CHECKS    →   24h INTERRUPT?   →   COMMITMENT
 ┌────────┐    ┌─────────────┐  ┌────────────────┐  ┌──────────────┐   ┌──────────────┐
 │ Search │    │ One-sentence│  │ Differentiation│  │ Enthusiasm   │   │ Public doc + │
 │ prior  │───▶│ guess +     │─▶│ Audience-fit   │─▶│ high? Wait   │──▶│ kill criteria│
 │ tries  │    │ confidence% │  │ Acquisition    │  │ 24h, restart │   │ D14/30/60/90 │
 └────────┘    └─────────────┘  │ Capacity       │  └──────────────┘   └──────┬───────┘
                                │ 7-Factor Wedge │                            │
                                └────────────────┘                            │
                                                                              ▼
                                                                       GO  → spec-driven-dev
                                                                       NO-GO → journal entry

Stage 0 before any /spec. Five framework gates, mandatory memory check against prior attempts, 24h pattern-interrupt for high-enthusiasm signals, public commitment artifact with pre-mortemed kill criteria. See skills/go-no-go/.

---

The spec-driven-dev loop

   DEFINE         PLAN           BUILD          TEST          REVIEW          SHIP
 ┌────────┐   ┌────────┐   ┌─────────┐   ┌────────┐   ┌────────┐   ┌────────┐
 │  Spec  │──▶│  Plan  │──▶│  Build  │──▶│  Test  │──▶│ Review │──▶│  Ship  │
 │  PRD   │   │ Tasks  │   │  Impl   │   │ Verify │   │  Gate  │   │  Tag   │
 └────────┘   └────────┘   └─────────┘   └────────┘   └────────┘   └────────┘
     ▲                                                                 │
     └─────────────────── feedback / regression ──────────────────────┘

One skill, six phases, explicit exit criteria per step. Pairs with go-no-go as Stage 1 of the pipeline. See skills/spec-driven-dev/.

---

The debug-hypothesis loop

                          ┌─────────────────────────────┐
                          │    all hypotheses rejected?  │
                          │    back to OBSERVE           │
                          └──────────┬──────────────────┘
                                     │
  OBSERVE  ──▶  HYPOTHESIZE  ──▶  EXPERIMENT  ──▶  CONCLUDE
     │              │                  │               │
     ▼              ▼                  ▼               ▼
  Reproduce      List 3-5          One test,       Root cause
  the bug,       causes +          max 5 lines,    confirmed
  collect        evidence           falsify         → fix +
  symptoms       for each          don't confirm   regression test
     │              │                  │               │
     │              │           ┌──────┘               │
     │              │           │ rejected?            │
     │              │           ▼                      │
     │              │     next hypothesis              │
     │              │                                  │
     └──────────────┴──── everything in DEBUG.md ──────┘

Scientific-method debugging. Prevents the #1 AI failure: bulldozing through a wrong theory. See skills/debug-hypothesis/.

---

The aggregate-N-sources loop

                    Trajectories-as-environment
            ╔════════════════════════════════════════╗
            ║   src_1   src_2   src_3  ...   src_N   ║
            ║   [..]    [..]    [..]         [..]    ║
            ║                                        ║
            ║   not concatenated. not summarized.    ║
            ║              navigated.                ║
            ╚═══════════════════╤════════════════════╝
                                ▼
            ┌────────────────────────────────────────┐
            │       AGGREGATOR  (lite agent)         │
            │   inspect_file / inspect_section       │
            │   search_sources / cross_pack_check    │
            │                                        │
            │   notes = []  (every claim → path:line)│
            │   budget = 25 → loop until coverage    │
            └───────────────────┬────────────────────┘
                                ▼
                  pack/  brief.md / findings.md /
                         sources.tsv / _aggregation_log.md

Agentic aggregation for long-horizon research. N raw notes → 1 structured pack with full path:line provenance + cross-source contradictions surfaced. See skills/wiki-aggregate/.

---

The build-until-pass loop

        ┌──────────────────────────────────────────────┐
        │  attempt > MAX (default 10)?                  │
        │  → STOP, report what's left, hand to human    │
        └───────────────────────┬──────────────────────┘
                                │ no
   RUN CHECK ──▶ exit 0? ──yes──▶  DONE  (green = exit code, not opinion)
       ▲             │
       │             │ no (red)
       │             ▼
       │      READ first error  ──▶  FIX smallest  ──▶  re-RUN
       │      verbatim, file:line    minimal diff,        │
       │                             never fake green     │
       └─────────────────────────────────────────────────┘
              one round = one fix + one re-run + one-line report

Turns the human out of the run → read error → fix one line → run again loop, and caps the agent so it can't bulldoze 200-line speculative rewrites between runs. Refuses to fake green (@ts-ignore, deleting tests, || true). See skills/build-until-pass/.

---

Skills

Skill	What it does
skills-map	The router — start here. Maps every skill in this collection to its phase of work (Decide → Spec → Build → Debug → Fan-out → Research → Media → Analyze) and walks a decision tree to pick the right one. Closes the discovery gap: a skill only fires if the agent remembers it exists and reaches for it at the right moment. Hand-curated index + typical pipelines + anti-rationalization table. Run on /skills-map or any "which skill should I use" doubt.
go-no-go	Stage 0 gate — NO-GO is the default. Runs before /spec to decide whether a project should start at all. Memory check against prior attempts + 5 framework gates (Differentiation · Audience–Market Fit · Acquisition Channel · Capacity · 7-Factor Wedge) + 24h pattern-interrupt if enthusiasm-high + public commitment doc with D14/D30/D60/D90 kill criteria. 3 starter packs (solo-founder · indie-dev · content-creator).
spec-driven-dev	Full SDLC workflow: Spec → Plan → Build → Test → Review → Ship. Anti-rationalization tables, verification gates, atomic commits. Pairs with go-no-go as Stage 1 of the pipeline.
debug-hypothesis	Scientific-method debugging: Observe → Hypothesize → Experiment → Conclude. Anti-bulldozer rules, max 5-line experiments, mandatory DEBUG.md evidence trail.
wiki-aggregate	Lift N≥3 raw research artifacts into one structured pack via agentic aggregation. Cheap-pass + tool-budgeted aggregator loop, every claim has path:line provenance, cross-source contradictions logged.
tavily-search	Web search + content extraction via the Tavily API. Use for fact-checking, docs lookup, source-cited research.
nano-banana	Text-to-image and image editing via Google's Nano Banana 2 (gemini-3.1-flash-image-preview). Supports 512 / 1K / 2K / 4K.
frontend-design	Build distinctive, production-grade frontend interfaces — bold aesthetic direction, intentional typography, and motion that avoids generic AI-slop UI. Adapted from Anthropic's official frontend-design skill (Apache-2.0).
google-analytics	Analyze GA4 data + ship an out-of-the-box SEO daily report — point it at one property and get organic KPIs (28d-vs-prior-28d), same-weekday anomaly detection, top organic landing pages, by-engine/country/device cuts, and prioritized recommendations. Plus general analyses (overview / sources / content / devices / seo). TypeScript scripts (Bun / npx tsx) over the official @google-analytics/data client, conversions↔keyEvents auto-detect. Reads GOOGLE_ANALYTICS_PROPERTY_ID + service-account JSON from env.
subagent-brief	Pre-flight discipline for spawning subagents. Anthropic does NOT share prefix across subagents — each one cold-starts on the full prompt. Compress every subagent prompt into a ≤200-word brief before spawning. Five rules + brief template + anti-rationalization table. Backed by arXiv 2604.25899 (Pythia, 2026).
build-until-pass	Bounded self-correcting loop: run the build/typecheck/test command → read the first error → apply the smallest fix → re-run → repeat until exit code 0 or a hard attempt cap (default 10). Green is the exit code, not the agent's opinion. Minimal-diff-per-round + cap kill both failure modes — human-as-while-loop and agent-as-bulldozer. Refuses to fake green (@ts-ignore, deleting tests, || true).

All skills read credentials from environment variables (TAVILY_API_KEY, GEMINI_API_KEY, etc.) — never hardcoded.

---

Quick install

Inside a running Claude Code session:

/plugin marketplace add LichAmnesia/lich-skills
/plugin install lich-skills@lich-skills

Done. All skills become available immediately. Verify:

/skills

# 1. Install Claude Code (if not already)
curl -fsSL https://claude.ai/install.sh | bash
# or: brew install --cask claude-code

# 2. Clone into the global skills directory
git clone https://github.com/LichAmnesia/lich-skills.git ~/.claude/skills/lich-skills

# 3. Start Claude Code
claude
> /skills

Full guide: docs/claude-code-setup.md.

gemini extensions install https://github.com/LichAmnesia/lich-skills

This repo ships a gemini-extension.json at the root, so Gemini CLI installs it as a first-class extension and auto-discovers every skills/*/SKILL.md. Verify:

gemini extensions list

Manual clone fallback:

npm install -g @google/gemini-cli
git clone https://github.com/LichAmnesia/lich-skills.git ~/.gemini/extensions/lich-skills

Full guide: docs/gemini-cli-setup.md.

# 1. Install Codex CLI
npm install -g @openai/codex
# or: brew install --cask codex

# 2. Install the skill collection
mkdir -p ~/.codex/skills
git clone https://github.com/LichAmnesia/lich-skills.git ~/.codex/skills/lich-skills

Full guide: docs/codex-setup.md.

---

Why the one-liners work