Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/adit-jain-srm/skill-forge/claude-mdgit clone --depth 1 https://github.com/Adit-Jain-srm/skill-forgeWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/adit-jain-srm/skill-forge/claude-md)<a href="https://agentmods.dev/instructions/adit-jain-srm/skill-forge/claude-md"><img src="https://agentmods.dev/badge/instructions/adit-jain-srm/skill-forge/claude-md.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00450 | $0.00450 |
| Opus 5 | $0.00225 | $0.00225 |
| Sonnet 5 | $0.00090 | $0.00090 |
| Haiku 4.5 | $0.00045 | $0.00045 |
Grade A, and why
skill-forge CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
CLAUDE.md — skill-forge
Read CONTEXT.md for project vocabulary. Read AGENTS.md for available skills and scripts.
Identity
This is a self-improving skill collection with two objectives: (1) make itself better each cycle, (2) produce skills that earn stars. The meta-agent in root SKILL.md orchestrates discovery, creation, routing, and improvement.
Key Facts
- Skills are in
skills/directory, each with a SKILL.md - Memory/state persists in
memory/*.json(indexed viascripts/index-memory.js) - Quality validation:
node scripts/validate-skill.js skills/<name>— must pass with 0 warnings - Self-check runs after every action (Phase 5 in root SKILL.md)
- Self-improvement:
node scripts/skillopt.jstracks outcomes and proposes prompt deltas - Routing:
node scripts/route-task.js "task"— TF-IDF semantic matching with compound skills - Publishing target: GitHub under Adit-Jain-srm org
- Install:
npx skills@latest add Adit-Jain-srm/skill-forge
When Modifying
- Never ship a skill without validation passing (0 warnings)
- Run
node tests/run-tests.jsafter any structural change (must pass all 98 tests) - Always update CONTEXT.md if adding new domain terms
- Learnings go in
memory/learnings.jsonand must change BEHAVIOR not just exist - After using a skill, record outcome:
node scripts/skillopt.js --record --skill <name> --task "..." --outcome <good|poor|mixed> - README frames skills as PROBLEMS not features
Reasoning Approach
When making decisions about skill creation, routing, or improvement:
- State the specific problem being solved (not vague goals)
- Consider what already exists (check memory/learnings.json and indexed-learnings.json)
- Validate against the best (not "good enough" — compare to top performers)
- Prove the result works (run tests, show evidence, never claim without verification)
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago First seen · 36 lines · 450 tokens per session scan A dc87ab75e683
skill-forge CLAUDE.md is an instructions file published in the GitHub repository Adit-Jain-srm/skill-forge (2 stars, last pushed 2mo ago), licensed MIT. It adds 450 tokens to every session, about $0.0022 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
crossframe-skill AGENTS.md
Instructions for xi-kari/crossframe-skill, covering crossframe / agent adapter, 仓库怎么理解, 显式调用边界, 显式调用 crossframe essay and 显式调用其它平行 skill.
crossframe-skill copilot-instructions.md
Instructions for xi-kari/crossframe-skill, a project described as: CrossFrame Skill Suite：面向 AI agent 的中文结构诊断、公共议题分析、组织复盘、历史材料处理、追问和成文工作流；旗舰 CrossFrame ProMax 提供 v8-only、精确点名、可对账可修复的独立结构推演 runtime。.
seo-geo-optimizer copilot-instructions.md
Instructions for Aryanpanwar10005/seo-geo-optimizer, covering seo & geo optimizer — universal ai skill, how to use this skill, part 1 — iron laws, part 2 — seo & geo execution phases and phase 0 — initialize.
ui-ux-design-pro-skill CLAUDE.md
Instructions for saifyxpro/ui-ux-design-pro-skill, covering project overview, cli commands, full design system generation (markdown output), search across all databases (sub-50ms) and audit ui code quality.
anki-card-skill GEMINI.md
Gemini CLI instructions for gong1414/anki-card-skill, a project described as: AI-powered Anki flashcard generator — works as a skill/plugin for Claude Code, Cursor, Copilot CLI, Gemini CLI, Codex & OpenCode. Exports to .tsv/.apkg with HTML formatting, hierarchical tags, and CJK support.
crossframe-skill CLAUDE.md
Instructions for xi-kari/crossframe-skill, a project described as: CrossFrame Skill Suite：面向 AI agent 的中文结构诊断、公共议题分析、组织复盘、历史材料处理、追问和成文工作流；旗舰 CrossFrame ProMax 提供 v8-only、精确点名、可对账可修复的独立结构推演 runtime。.