thkt/dotclaude
Personal Claude Code configuration — 28 skills, 28 agents, deterministic multi-agent workflows, and quality-gate hooks
field: AI / ML workflow: PR / code reviewFrontend & designPlanning & specsSecurity reviewTesting & TDD runs claude in CI
ARCHETYPE
Fort Knox
Deny lists, pre-tool hooks, sandboxing. Nothing touches prod without a signature.
GUARDRAILS
4/5
Blocks destructive commands · Protects secrets · No YOLO mode · Sandbox or ask-first rules · details
Quality 42/100 Guardrails v2
Copy this rig
# review before running: this installs third-party code $ npx degit thkt/dotclaude/agents ./rig-dotclaude/agents $ npx degit thkt/dotclaude/skills ./rig-dotclaude/skills
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 thkt/dotclaude/.ja/skills/ablate .claude/skills/ablate $ npx degit thkt/dotclaude/.ja/skills/census .claude/skills/census $ npx degit thkt/dotclaude/.ja/skills/challenge .claude/skills/challenge $ npx degit thkt/dotclaude/.ja/skills/checkout .claude/skills/checkout $ npx degit thkt/dotclaude/.ja/skills/commit .claude/skills/commit $ npx degit thkt/dotclaude/.ja/skills/dr .claude/skills/dr $ npx degit thkt/dotclaude/.ja/skills/fix .claude/skills/fix $ npx degit thkt/dotclaude/.ja/skills/implement .claude/skills/implement $ npx degit thkt/dotclaude/.ja/skills/issue .claude/skills/issue $ npx degit thkt/dotclaude/.ja/skills/outcome .claude/skills/outcome $ npx degit thkt/dotclaude/.ja/skills/pr .claude/skills/pr $ npx degit thkt/dotclaude/.ja/skills/preview .claude/skills/preview $ npx degit thkt/dotclaude/.ja/skills/qualify .claude/skills/qualify $ npx degit thkt/dotclaude/.ja/skills/research .claude/skills/research $ npx degit thkt/dotclaude/.ja/skills/scoping .claude/skills/scoping $ npx degit thkt/dotclaude/.ja/skills/scribe .claude/skills/scribe $ npx degit thkt/dotclaude/.ja/skills/slice .claude/skills/slice $ npx degit thkt/dotclaude/.ja/skills/think .claude/skills/think $ npx degit thkt/dotclaude/.ja/skills/transcribe .claude/skills/transcribe $ npx degit thkt/dotclaude/.ja/skills/use-cli-ast-grep .claude/skills/use-cli-ast-grep $ npx degit thkt/dotclaude/.ja/skills/use-cli-codegraph .claude/skills/use-cli-codegraph $ npx degit thkt/dotclaude/.ja/skills/use-cli-gcloud .claude/skills/use-cli-gcloud $ npx degit thkt/dotclaude/.ja/skills/use-cli-recall .claude/skills/use-cli-recall $ npx degit thkt/dotclaude/.ja/skills/use-cli-scout .claude/skills/use-cli-scout $ npx degit thkt/dotclaude/.ja/skills/use-context-reviewer-readability .claude/skills/use-context-reviewer-readability $ npx degit thkt/dotclaude/.ja/skills/use-context-reviewer-security .claude/skills/use-context-reviewer-security $ npx degit thkt/dotclaude/.ja/skills/use-context-reviewer-silence .claude/skills/use-context-reviewer-silence $ npx degit thkt/dotclaude/.ja/skills/use-context-reviewer-testability .claude/skills/use-context-reviewer-testability $ npx degit thkt/dotclaude/.ja/skills/use-context-root-cause-analysis .claude/skills/use-context-root-cause-analysis $ npx degit thkt/dotclaude/.ja/skills/use-workflow-pageshot .claude/skills/use-workflow-pageshot $ npx degit thkt/dotclaude/.ja/skills/use-workflow-tdd-cycle .claude/skills/use-workflow-tdd-cycle
$ curl -fsSL --create-dirs -o .claude/agents/A11Y.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/calibration/A11Y.md $ curl -fsSL --create-dirs -o .claude/agents/CHX.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/calibration/CHX.md $ curl -fsSL --create-dirs -o .claude/agents/CQ.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/calibration/CQ.md $ curl -fsSL --create-dirs -o .claude/agents/DP.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/calibration/DP.md $ curl -fsSL --create-dirs -o .claude/agents/DRY.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/calibration/DRY.md $ curl -fsSL --create-dirs -o .claude/agents/EFF.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/calibration/EFF.md $ curl -fsSL --create-dirs -o .claude/agents/OPS.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/calibration/OPS.md $ curl -fsSL --create-dirs -o .claude/agents/PE.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/calibration/PE.md $ curl -fsSL --create-dirs -o .claude/agents/PQ.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/calibration/PQ.md $ curl -fsSL --create-dirs -o .claude/agents/RC.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/calibration/RC.md $ curl -fsSL --create-dirs -o .claude/agents/REUSE.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/calibration/REUSE.md $ curl -fsSL --create-dirs -o .claude/agents/RP.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/calibration/RP.md $ curl -fsSL --create-dirs -o .claude/agents/SEC.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/calibration/SEC.md $ curl -fsSL --create-dirs -o .claude/agents/SF.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/calibration/SF.md $ curl -fsSL --create-dirs -o .claude/agents/TC.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/calibration/TC.md $ curl -fsSL --create-dirs -o .claude/agents/TEST.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/calibration/TEST.md $ curl -fsSL --create-dirs -o .claude/agents/evidence-report-template.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/evidence-report-template.md $ curl -fsSL --create-dirs -o .claude/agents/finding-disposition.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/finding-disposition.md $ curl -fsSL --create-dirs -o .claude/agents/finding-registry.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/finding-registry.md $ curl -fsSL --create-dirs -o .claude/agents/finding-schema.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/finding-schema.md $ curl -fsSL --create-dirs -o .claude/agents/prompt-quality-checks.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/prompt-quality-checks.md $ curl -fsSL --create-dirs -o .claude/agents/root-cause-synthesis.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/root-cause-synthesis.md $ curl -fsSL --create-dirs -o .claude/agents/simplification-rules.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/_lib/simplification-rules.md $ curl -fsSL --create-dirs -o .claude/agents/critic-audit.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/critics/critic-audit.md $ curl -fsSL --create-dirs -o .claude/agents/critic-design.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/critics/critic-design.md $ curl -fsSL --create-dirs -o .claude/agents/critic-evidence.md https://raw.githubusercontent.com/thkt/dotclaude/main/agents/critics/critic-evidence.md $ curl -fsSL --create-dirs -o .claude/agents/enhancer-code.md https://raw.githubusercontent.<redacted>-code.md $ curl -fsSL --create-dirs -o .claude/agents/enhancer-evidence.md https://raw.githubusercontent.<redacted>-evidence.md $ curl -fsSL --create-dirs -o .claude/agents/enhancer-integration.md https://raw.githubusercontent.<redacted>-integration.md $ curl -fsSL --create-dirs -o .claude/agents/explorer-feature.md https://raw.githubusercontent.<redacted>-feature.md $ curl -fsSL --create-dirs -o .claude/agents/generator-snapshot.md https://raw.githubusercontent.<redacted>-snapshot.md $ curl -fsSL --create-dirs -o .claude/agents/generator-test.md https://raw.githubusercontent.<redacted>-test.md $ curl -fsSL --create-dirs -o .claude/agents/resolver-build.md https://raw.githubusercontent.<redacted>-build.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer-accessibility.md https://raw.githubusercontent.<redacted>-accessibility.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer-causation.md https://raw.githubusercontent.<redacted>-causation.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer-conformance.md https://raw.githubusercontent.<redacted>-conformance.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer-coverage.md https://raw.githubusercontent.<redacted>-coverage.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer-design.md https://raw.githubusercontent.<redacted>-design.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer-duplication.md https://raw.githubusercontent.<redacted>-duplication.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer-efficiency.md https://raw.githubusercontent.<redacted>-efficiency.md
Merge into .claude/settings.json (project) or ~/.claude/settings.json (user). Hook commands reference scripts in the source repo — copy those too.
{
"permissions": {
"deny": [
"Bash(git stash clear *)",
"Bash(git stash drop *)",
"Bash(*--exec-path*)",
"Bash(*--no-verify*)",
"Bash(*--receive-pack*)",
"Bash(*--upload-pack*)",
"Bash(curl *)",
"Bash(wget *)",
"Bash(kill -9 *)",
"Bash(sudo *)",
"Bash(rm *)",
"Bash(rmdir *)",
"Bash(unlink *)",
"Bash(shred *)",
"Bash(find * -delete*)",
"Read(.env)",
"Read(.env.*)",
"Read(**/.env)",
"Read(**/.env.*)",
"Read(~/.ssh/id_rsa)",
"Read(~/.ssh/id_ed25519)",
"Read(**/*token*.db)",
"Read(**/.token*)",
"Read(**/*_token*)",
"Read(**/token.json)",
"Read(**/*.key)",
"Read(**/*.pem)",
"Read(**/*_key)",
"Read(~/.aws/**)",
"Read(~/.kube/**)",
"Read(~/.zshenv)",
"Read(**/*credentials*)",
"Read(**/*secret*)",
"Read(**/secrets/**)",
"Edit(.env*)",
"Edit(**/.env)",
"Edit(**/.env.*)",
"Edit(**/secrets/**)",
"Edit(~/.zshrc)",
"Edit(~/.zshenv)",
"Edit(~/.bashrc)",
"Edit(~/.bash_profile)",
"Edit(~/.zprofile)",
"Edit(~/.gitconfig)",
"Edit(~/.gnupg/**)",
"Edit(~/.ssh/**)",
"Edit(~/.aws/**)",
"Edit(~/.kube/**)",
"Edit(~/Library/LaunchAgents/**)",
"NotebookEdit"
],
"ask": [
"Bash(git -c *)",
"Bash(git branch -D *)",
"Bash(git checkout -- *)",
"Bash(git clean *)",
"Bash(git rebase *)",
"Bash(git reset *)",
"Bash(git restore *)",
"Bash(git rm *)",
"Edit(~/.claude/settings.json)",
"Edit(~/.claude/hooks/**)",
"Edit(~/.claude/rules/**)",
"Edit(//**/.claude/agent-memory/**)",
"Edit(//**/.claude/agent-memory-local/**)"
]
},
"sandbox": {
"enabled": true
}
} Skills (31)
ablatecensuschallengecheckoutcommitdrfiximplementissueoutcomeprpreviewqualifyresearchscopingscribeslicethinktranscribeuse-cli-ast-grepuse-cli-codegraphuse-cli-gclouduse-cli-recalluse-cli-scoutuse-context-reviewer-readabilityuse-context-reviewer-securityuse-context-reviewer-silenceuse-context-reviewer-testabilityuse-context-root-cause-analysisuse-workflow-pageshotuse-workflow-tdd-cycle
Subagents (51)
| A11Y | — |
| CHX | — |
| CQ | — |
| DP | — |
| DRY | — |
| EFF | — |
| OPS | — |
| PE | — |
| PQ | — |
| RC | — |
| REUSE | — |
| RP | — |
| SEC | — |
| SF | — |
| TC | — |
| TEST | — |
| evidence-report-template | — |
| finding-disposition | — |
| finding-registry | — |
| finding-schema | — |
| prompt-quality-checks | — |
| root-cause-synthesis | — |
| simplification-rules | — |
| critic-audit model: opus | Delegate after a review pass, to challenge each finding for false positives and return a verdict per finding. |
| critic-design model: opus | Delegate before a design proposal is adopted, to attack it for hidden assumptions, costs, and failure modes. |
| critic-evidence model: opus | Delegate after a review pass, to verify each finding by tracing a concrete execution path. The verifier that complements critic-audit, the challenger. |
| enhancer-code model: opus | Delegate after a change lands, to strip AI slop, redundant tests, and defensive excess from the changed files without changing behavior. |
| enhancer-evidence model: opus | Delegate at the end of an assert run, to reconcile static findings, dynamic evidence, and adversarial results into issues, root causes, and a report. |
| enhancer-integration model: opus | Delegate after the challenge triage of an audit, to merge the survivors into cross-domain root causes. |
| explorer-feature model: opus | Delegate when research needs the shape of a feature, to trace its execution paths, map its architecture, and list the files to read. |
| generator-snapshot model: sonnet | Use once per audit run, after the findings are final, to persist the snapshot record. Writes the payload verbatim to a temp file and runs snapshot.ts once. Does not review code or judge findings. |
| generator-test model: opus | Use from a bug fix, before touching the cause, to generate the failing test that reproduces the reported symptom. Does not implement code. |
| resolver-build model: opus | Use when a build or type-check command reports errors, to fix them with the smallest diff. No refactoring or architectural change. |
| reviewer-accessibility model: opus | Delegate when a diff touches HTML, CSS, or UI components, to check WCAG 2.2 compliance. |
| reviewer-causation model: opus | Delegate when a diff fixes a bug or adds a workaround, to check whether the change removes the cause or silences the symptom. |
| reviewer-conformance model: opus | Delegate after an implementation lands, to check the diff against the originating issue or spec. Reports missing, scope_creep, and wrong findings, each with the spec line quoted. |
| reviewer-coverage model: opus | Delegate when a diff touches tests or testable logic, to find behavior left untested and tests that pass even when the implementation breaks. |
| reviewer-design model: opus | Delegate when a diff adds or reshapes a module boundary (function, class, hook, package), to check whether each module earns its interface by the deletion test. |
| reviewer-duplication model: opus | Delegate when a diff or directory may repeat logic across 2 or more files, to find the duplication and propose one shared utility. |
| reviewer-efficiency model: sonnet | Delegate when a diff touches loops, request handlers, I/O, or concurrency, to find work the code does more than once or more than needed. |
Plugins (7)
agent-browser@agent-browserrust-analyzer-lsp@claude-plugins-officialtypescript-lsp@claude-plugins-officialhtml@skillstypesafe@typesafe-airedact@thkt-modsfigma@claude-plugins-official
Permissions
deny (50)
Bash(git stash clear *)
Bash(git stash drop *)
Bash(*--exec-path*)
Bash(*--no-verify*)
Bash(*--receive-pack*)
Bash(*--upload-pack*)
Bash(curl *)
Bash(wget *)
Bash(kill -9 *)
Bash(sudo *)
Bash(rm *)
Bash(rmdir *)
Bash(unlink *)
Bash(shred *)
Bash(find * -delete*)
Read(.env)
Read(.env.*)
Read(**/.env)
Read(**/.env.*)
Read(~/.ssh/id_rsa)
Read(~/.ssh/id_ed25519)
Read(**/*token*.db)
Read(**/.token*)
Read(**/*_token*)
Read(**/token.json)
Read(**/*.key)
Read(**/*.pem)
Read(**/*_key)
Read(~/.aws/**)
Read(~/.kube/**)
Read(~/.zshenv)
Read(**/*credentials*)
Read(**/*secret*)
Read(**/secrets/**)
Edit(.env*)
Edit(**/.env)
Edit(**/.env.*)
Edit(**/secrets/**)
Edit(~/.zshrc)
Edit(~/.zshenv)
Edit(~/.bashrc)
Edit(~/.bash_profile)
Edit(~/.zprofile)
Edit(~/.gitconfig)
Edit(~/.gnupg/**)
Edit(~/.ssh/**)
Edit(~/.aws/**)
Edit(~/.kube/**)
Edit(~/Library/LaunchAgents/**)
NotebookEdit
ask (13)
Bash(git -c *)
Bash(git branch -D *)
Bash(git checkout -- *)
Bash(git clean *)
Bash(git rebase *)
Bash(git reset *)
Bash(git restore *)
Bash(git rm *)
Edit(~/.claude/settings.json)
Edit(~/.claude/hooks/**)
Edit(~/.claude/rules/**)
Edit(//**/.claude/agent-memory/**)
Edit(//**/.claude/agent-memory-local/**)
allow (34)
ExitPlanMode(*)
LS(*)
ListMcpResourcesTool(*)
ReadMcpResourceTool(*)
TaskStop(*)
Bash(gh *)
Bash(git *)
Bash(claude --print *)
Bash(bun *)
Bash(npm *)
Bash(pnpm *)
Bash(yarn *)
Bash(na *)
Bash(nci *)
Bash(ni *)
Bash(nlx *)
Bash(nr *)
Bash(nun *)
Bash(nup *)
Bash(kubectl describe *)
Bash(kubectl get *)
Bash(agent-browser *)
Bash(agy *)
Bash(ast-grep *)
Bash(bfs *)
Bash(cargo *)
Bash(codex review *)
Bash(recall *)
Bash(sae get *)
Bash(sae search *)
Bash(sae status *)
Bash(scout *)
Bash(ugrep *)
Bash(xr *)
Similar rigs
martinemde/dotfiles
✨ AI-first dotfiles with chezmoi
Orchestrator 3.1k tok · · Q18
orin-dx/wisp-plugins
Ten plugins for coding agents, one per stage of the development lifecycle, joined by shared JSON schemas and a common verification gate. For Claude Code, Cursor and Antigravity.
Skill Collector 3.2k tok · · Q30
ivy/dotfiles
Agent-managed developer environment. I describe what I want, they do the rest.
Skill Collector 3.2k tok · · Q15
damtzi/dotfiles
My personal dotfiles.
Skill Collector 1.5k tok · · Q0