~ / rigs / Ces107 / agent-forge

Ces107/agent-forge

Free agent skills and Claude Code plugins: launch-check (pre-launch audit of any URL: exposed keys, open database, SEO, phone design) and done-gate (no "done" until the tests pass). SKILL.md format for Claude Code, Codex, Cursor and Copilot

↗ GitHub ★ 0 MIT updated 25d ago project Claude Code Claude plugin
stack: DockerPython
share on X
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
CONTEXT TAX · EVERY TURN
~889 tokens
Featherweight · median rig: 2.3k · breakdown
GUARDRAILS
0/5
No committed guardrails · details
Quality 0/100 Guardrails v2

Copy this rig

# review before running: this installs third-party code
$ npx degit Ces107/agent-forge/.claude ./rig-agent-forge  # inspect, then merge into .claude/

MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.

$ npx degit Ces107/agent-forge/plugins/agent-forge/skills/adversarial-review-gate .claude/skills/adversarial-review-gate
$ npx degit Ces107/agent-forge/plugins/agent-forge/skills/done-gate .claude/skills/done-gate
$ npx degit Ces107/agent-forge/plugins/agent-forge/skills/launch-check .claude/skills/launch-check
$ npx degit Ces107/agent-forge/plugins/agent-forge/skills/proven-code-review .claude/skills/proven-code-review
$ npx degit Ces107/agent-forge/plugins/agent-forge/skills/spec-trace-gate .claude/skills/spec-trace-gate
$ npx degit Ces107/agent-forge/plugins/agent-forge/skills/verify-gate .claude/skills/verify-gate
$ curl -fsSL --create-dirs -o .claude/agents/adversarial-reviewer.md https://raw.githubusercontent.com/Ces107/agent-forge/main/.claude/agents/adversarial-reviewer.md
$ curl -fsSL --create-dirs -o .claude/agents/architect.md https://raw.githubusercontent.com/Ces107/agent-forge/main/.claude/agents/architect.md
$ curl -fsSL --create-dirs -o .claude/agents/implementer.md https://raw.githubusercontent.com/Ces107/agent-forge/main/.claude/agents/implementer.md
$ curl -fsSL --create-dirs -o .claude/agents/planner.md https://raw.githubusercontent.com/Ces107/agent-forge/main/.claude/agents/planner.md
$ curl -fsSL --create-dirs -o .claude/agents/verifier.md https://raw.githubusercontent.com/Ces107/agent-forge/main/.claude/agents/verifier.md

This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 16.2k rigs:

{
  "permissions": {
    "deny": [
      "Read(./.env)",
      "Read(./.env.*)",
      "Read(**/.env)",
      "Read(~/.ssh/**)",
      "Bash(git push --force:*)",
      "Read(**/*.pem)",
      "Read(.env)",
      "Read(~/.aws/**)",
      "Read(**/*.key)",
      "Read(**/.env.*)",
      "Bash(git push --force*)",
      "Bash(rm -rf:*)",
      "Bash(rm -rf *)",
      "Bash(sudo *)",
      "Bash(sudo:*)",
      "Bash(rm -rf /)",
      "Bash(git push -f:*)",
      "Read(.env.*)",
      "Bash(git reset --hard*)",
      "Bash(rm -rf /*)"
    ],
    "ask": [
      "Bash(git push:*)",
      "Bash(git push *)",
      "Bash(gh pr merge *)",
      "Bash(git commit:*)",
      "Bash(rm *)",
      "Bash(gh pr merge:*)",
      "Bash(gh pr create *)",
      "Bash(git reset --hard:*)",
      "Bash(npm publish:*)",
      "Bash(git rebase *)"
    ]
  }
}

Skills (6)

Subagents (5)

adversarial-reviewer
model: sonnet
MUST BE USED as pipeline stage 4. Attacks the Implementer's output for correctness bugs — especially concurrency, idempotency and money-conservation defects. The gate HALTS unless this stage emits at
architect
model: opus
MUST BE USED as pipeline stage 2. Turns an acceptance spec into ADR-style decision records. Produces NO application code — only design decisions with explicit trade-offs and a contract the Implementer
implementer
model: sonnet
MUST BE USED as pipeline stage 3. Writes the application code and the first test layer for the inner project, honouring the Architect's ADR contract. Returns artifact paths + a one-line summary, never
planner
model: sonnet
MUST BE USED as pipeline stage 3 (Tasks), after the Architect. Decomposes the ADR contract into a flat list of implementation tasks, each traceable backward to a spec acceptance criterion (AC-N) and f
verifier
model: haiku
MUST BE USED as the final pipeline gate. Runs the actual test suite, coverage, type-check, lint and security scan and refuses to let the Commit stage proceed unless every check exits 0. Enforces the a

Similar rigs

copied ✓