Eric-Lingren/claude-code-toolkit
My Claude Code toolkit: a skill and agent pipeline from plan to merged PR, with evals, model tiering and guardrail hooks
stack: Python
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
Copy this rig
# review before running: this installs third-party code $ npx degit Eric-Lingren/claude-code-toolkit/agents ./rig-claude-code-toolkit/agents $ npx degit Eric-Lingren/claude-code-toolkit/skills ./rig-claude-code-toolkit/skills $ npx degit Eric-Lingren/claude-code-toolkit/hooks ./rig-claude-code-toolkit/hooks
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 Eric-Lingren/claude-code-toolkit/skills/build-code .claude/skills/build-code $ npx degit Eric-Lingren/claude-code-toolkit/skills/cc-usage-analytics .claude/skills/cc-usage-analytics $ npx degit Eric-Lingren/claude-code-toolkit/skills/ci-fix .claude/skills/ci-fix $ npx degit Eric-Lingren/claude-code-toolkit/skills/clean-scaffolding .claude/skills/clean-scaffolding $ npx degit Eric-Lingren/claude-code-toolkit/skills/debug .claude/skills/debug $ npx degit Eric-Lingren/claude-code-toolkit/skills/dispatch-tasks .claude/skills/dispatch-tasks $ npx degit Eric-Lingren/claude-code-toolkit/skills/find-work .claude/skills/find-work $ npx degit Eric-Lingren/claude-code-toolkit/skills/grill-me .claude/skills/grill-me $ npx degit Eric-Lingren/claude-code-toolkit/skills/grill-with-docs .claude/skills/grill-with-docs $ npx degit Eric-Lingren/claude-code-toolkit/skills/handoff .claude/skills/handoff $ npx degit Eric-Lingren/claude-code-toolkit/skills/improve-agent-benchmarks .claude/skills/improve-agent-benchmarks $ npx degit Eric-Lingren/claude-code-toolkit/skills/improve-codebase-architecture .claude/skills/improve-codebase-architecture $ npx degit Eric-Lingren/claude-code-toolkit/skills/improve-component .claude/skills/improve-component $ npx degit Eric-Lingren/claude-code-toolkit/skills/improve-directory-structure .claude/skills/improve-directory-structure $ npx degit Eric-Lingren/claude-code-toolkit/skills/improve-skill-benchmarks .claude/skills/improve-skill-benchmarks $ npx degit Eric-Lingren/claude-code-toolkit/skills/improve-skill-learnings .claude/skills/improve-skill-learnings $ npx degit Eric-Lingren/claude-code-toolkit/skills/investigate .claude/skills/investigate $ npx degit Eric-Lingren/claude-code-toolkit/skills/offload .claude/skills/offload $ npx degit Eric-Lingren/claude-code-toolkit/skills/pick .claude/skills/pick $ npx degit Eric-Lingren/claude-code-toolkit/skills/pr-code-review .claude/skills/pr-code-review $ npx degit Eric-Lingren/claude-code-toolkit/skills/pr-revise .claude/skills/pr-revise $ npx degit Eric-Lingren/claude-code-toolkit/skills/prototype .claude/skills/prototype $ npx degit Eric-Lingren/claude-code-toolkit/skills/register-skill .claude/skills/register-skill $ npx degit Eric-Lingren/claude-code-toolkit/skills/relay .claude/skills/relay $ npx degit Eric-Lingren/claude-code-toolkit/skills/retro .claude/skills/retro $ npx degit Eric-Lingren/claude-code-toolkit/skills/run-task-followups .claude/skills/run-task-followups $ npx degit Eric-Lingren/claude-code-toolkit/skills/sprout-seed .claude/skills/sprout-seed $ npx degit Eric-Lingren/claude-code-toolkit/skills/tasks-to-linear .claude/skills/tasks-to-linear $ npx degit Eric-Lingren/claude-code-toolkit/skills/tdd .claude/skills/tdd $ npx degit Eric-Lingren/claude-code-toolkit/skills/tldr-tech .claude/skills/tldr-tech $ npx degit Eric-Lingren/claude-code-toolkit/skills/to-e2e-tasks .claude/skills/to-e2e-tasks $ npx degit Eric-Lingren/claude-code-toolkit/skills/to-seed .claude/skills/to-seed $ npx degit Eric-Lingren/claude-code-toolkit/skills/to-spec .claude/skills/to-spec $ npx degit Eric-Lingren/claude-code-toolkit/skills/to-tasks .claude/skills/to-tasks $ npx degit Eric-Lingren/claude-code-toolkit/skills/worktree .claude/skills/worktree
$ curl -fsSL --create-dirs -o .claude/agents/answer-composer.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/egress/answer-composer.md $ curl -fsSL --create-dirs -o .claude/agents/export-tasks.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/egress/export-tasks.md $ curl -fsSL --create-dirs -o .claude/agents/export-tasks-github.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/egress/github/export-tasks-github.md $ curl -fsSL --create-dirs -o .claude/agents/post-github.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/egress/github/post-github.md $ curl -fsSL --create-dirs -o .claude/agents/export-tasks-linear.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/egress/linear/export-tasks-linear.md $ curl -fsSL --create-dirs -o .claude/agents/post-linear.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/egress/linear/post-linear.md $ curl -fsSL --create-dirs -o .claude/agents/export-tasks-notion.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/egress/notion/export-tasks-notion.md $ curl -fsSL --create-dirs -o .claude/agents/post-slack.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/egress/slack/post-slack.md $ curl -fsSL --create-dirs -o .claude/agents/investigator-code.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/investigate/investigator-code.md $ curl -fsSL --create-dirs -o .claude/agents/investigator-github.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/investigate/investigator-github.md $ curl -fsSL --create-dirs -o .claude/agents/investigator-linear.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/investigate/investigator-linear.md $ curl -fsSL --create-dirs -o .claude/agents/investigator-notion.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/investigate/investigator-notion.md $ curl -fsSL --create-dirs -o .claude/agents/investigator-web.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/investigate/investigator-web.md $ curl -fsSL --create-dirs -o .claude/agents/investigator.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/investigate/investigator.md $ curl -fsSL --create-dirs -o .claude/agents/artifact-grounding-judge.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/learning/artifact-grounding-judge.md $ curl -fsSL --create-dirs -o .claude/agents/attribution-tracer.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/learning/attribution-tracer.md $ curl -fsSL --create-dirs -o .claude/agents/capture-learning.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/learning/capture-learning.md $ curl -fsSL --create-dirs -o .claude/agents/pick-scorer.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/pick/pick-scorer.md $ curl -fsSL --create-dirs -o .claude/agents/persona-accuracy.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/seed-review/persona-accuracy.md $ curl -fsSL --create-dirs -o .claude/agents/persona-coherence.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/seed-review/persona-coherence.md $ curl -fsSL --create-dirs -o .claude/agents/persona-completeness.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/seed-review/persona-completeness.md $ curl -fsSL --create-dirs -o .claude/agents/persona-grounding.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/seed-review/persona-grounding.md $ curl -fsSL --create-dirs -o .claude/agents/persona-judge.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/seed-review/persona-judge.md $ curl -fsSL --create-dirs -o .claude/agents/architecture-auditor.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/shared/architecture-auditor.md $ curl -fsSL --create-dirs -o .claude/agents/context-loader.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/shared/context-loader.md $ curl -fsSL --create-dirs -o .claude/agents/build-code.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/skill-agents/build-code.md $ curl -fsSL --create-dirs -o .claude/agents/pr-code-review.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/skill-agents/pr-code-review.md $ curl -fsSL --create-dirs -o .claude/agents/to-tasks.md https://raw.githubusercontent.com/Eric-Lingren/claude-code-toolkit/main/agents/skill-agents/to-tasks.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 (35)
build-codecc-usage-analyticsci-fixclean-scaffoldingdebugdispatch-tasksfind-workgrill-megrill-with-docshandoffimprove-agent-benchmarksimprove-codebase-architectureimprove-componentimprove-directory-structureimprove-skill-benchmarksimprove-skill-learningsinvestigateoffloadpickpr-code-reviewpr-reviseprototyperegister-skillrelayretrorun-task-followupssprout-seedtasks-to-lineartddtldr-techto-e2e-tasksto-seedto-specto-tasksworktree
Subagents (28)
| answer-composer model: sonnet | > |
| export-tasks model: sonnet | Export triage tasks from a task file to configured external destinations. Resolves each item's (deliverable, domain) pair against task-routing.json, shows a dry-run table with wall-crossing warnings, |
| export-tasks-github model: haiku | GitHub Issues write adapter. Creates a single GitHub issue from structured JSON input. Spawned by the export-tasks coordinator per gh-issues item. Returns the created issue URL as the sole response li |
| post-github model: haiku | > |
| export-tasks-linear model: haiku | Linear write adapter. Creates a single Linear issue from structured JSON input using the Linear MCP tool. Spawned by the export-tasks coordinator per linear item. Returns the created issue URL as the |
| post-linear model: haiku | > |
| export-tasks-notion model: haiku | Notion write adapter. Creates a single Notion page from structured JSON input using MCP or API-token auth. Spawned by the export-tasks coordinator per notion item. Returns the created page URL as the |
| post-slack model: haiku | > |
| investigator-code model: sonnet | > |
| investigator-github model: sonnet | > |
| investigator-linear model: sonnet | > |
| investigator-notion model: sonnet | > |
| investigator-web model: sonnet | > |
| investigator model: opus | > |
| artifact-grounding-judge model: haiku | Artifact-evidence grounding judge. Receives a draft v2 attribution record from the attribution-tracer. Verifies each evidence entry against its cited artifact file — positive anchors (quote present) o |
| attribution-tracer model: sonnet | Attribution tracing agent. Receives a confirmed issue description and fix, walks the provenance chain backward to find the earliest escape point, classifies the cause, and drafts a v2 attribution lear |
| capture-learning model: sonnet | End-of-run learning capture agent. Receives a rough correction-event description plus candidate verbatim anchors from a skill, expands it into a schema-valid v2 self-learning entry, grounds it with th |
| pick-scorer model: sonnet | Scores up to 25 pre-sorted /pick candidates for effort and returns a per-candidate score plus a one-line reason as JSON. Pure function of its input; reads nothing else. Spawned by the pick skill. Cont |
| persona-accuracy model: haiku | Adversary persona that hunts semantic drift and stale-resolution in a draft seed. Spawned by to-seed verification stage. Returns refutations with cited transcript spans. |
| persona-coherence model: haiku | Adversary persona that hunts contradiction, misclassification, dropped human dispositions, and relabel-resurrection in a draft seed. Spawned by to-seed verification stage. Returns refutations with cit |
| persona-completeness model: haiku | Adversary persona that hunts missed branches, premature closure, dropped dependencies, and merge-loss in a draft seed. Spawned by to-seed verification stage. Returns refutations with cited transcript |
| persona-grounding model: haiku | Adversary persona that hunts unsupported content and fabricated rationale in a draft seed. Spawned by to-seed verification stage. Returns refutations with cited transcript spans. |
| persona-judge model: sonnet | Evidence-bound adjudicator. Receives a BATCH of refutations and a windowed evidence pack, returns one verdict per refutation. Spawned by to-seed verification stage — once as a round-1 screener, then 3 |
| architecture-auditor model: sonnet | Read-only agent that audits a target SKILL.md for architecture anti-patterns and agent extraction opportunities. Detects 8 signals (A1-A8) across three categories: context-blowup risk, extraction cand |
| context-loader model: haiku | Discovers and distills project context (CONTEXT.md, ADRs, and typed brand/design sources) into a structured JSON payload. Spawned by skills before they need domain vocabulary, architectural decisions, |
| build-code model: sonnet | Execute tasks from a tasks JSON file sequentially, one build-runner spawn per task. Handles branching, status updates, blocker detection, leaf/blocker failure policy, and deferred HITL tasks without h |
| pr-code-review model: sonnet | Review the current diff, or a PR number/branch/path target, for correctness bugs and reuse/simplification/efficiency cleanups at the given effort level. Spawnable by any skill via subagent_type pr-cod |
| to-tasks model: sonnet | Break a seed or PRD into independently-grabbable tasks using tracer-bullet vertical slices, then write a JSON task file to docs/tasks/. Use when user wants to convert a seed file or PRD into AI-ready |
Similar rigs
davner/dotfiles
My Macs as code: nix-darwin and home-manager for macOS settings, Homebrew, CLI tools and zsh, plus a Claude Code setup of skills, hooks and rules. One branch, keyed by macOS username.
Skill Collector 4.3k tok · · Q0
mokevnin/dotfiles
using vim
Skill Collector 1.2k tok · · Q0
Tahseenm/dotfiles
Automate Mac OS and Ubuntu machine setup
Skill Collector 1.3k tok · · Q0
ZenoWangzy/zenoforge
Self-hosted software factory: coding agents take GitHub issues through triage, implementation, review, CI repair and merge. Source-available (PolyForm Noncommercial).
Skill Collector 2.5k tok · · Q0