MCP threat-model artifact for a scaffolded harness. Reports allowed/denied tools, dangerous permissions count, secrets reachability, network/shell/file-write grants, default-deny posture. Verdict: clean (exit 0) / medium (exit 1) / high (exit 2). The 'enterprise gold' artifact for PR + compliance review.
インストールする前に、エージェントに与えられる指示の中身を確認できます。
Codex skill: MCP threat-model artifact for PR / compliance review (iter 112 → iter 114). User-labelled "enterprise gold."
Renders the existing mcp-scan findings as a clean threat-model artifact
in the shape a security / compliance reviewer wants to see attached to a
PR:
MCP Threat Model
Allowed tools: 3
Denied tools: 14
Dangerous permissions: 0
Secrets reachable: no
Network access: no
Shell access: no
File write: no
Default-deny policy: yes
Audit log: yes
Verdict: clean (exit 0)
Same underlying scan as mcp-scan, presented as a single-screen artifact.
| Verdict | Exit | Triggers |
|---|---|---|
clean | 0 | no dangerous perms, no secret exposure |
medium | 1 | network OR file-write granted, OR no audit log |
high | 2 | shell granted OR default-deny OFF OR secrets reachable |
The "secrets reachable" heuristic is conservative: true when MCP is in
use AND deny rules don't guard .env* AND allow rules include any
Read(...) grant.
/threat-model path=./my-harness
/threat-model path=./my-harness bundle=true
harness threat-model ./my-harness # text artifact
harness threat-model ./my-harness --json # full envelope
harness threat-model ./my-harness --bundle # ADR-031 schema-1
harness threat-model ./my-harness --out tm.json # write to file
.harness/mcp-policy.json or .claude/settings.json
permissionsThe artifact is small enough to paste verbatim into the PR description.
validate-harness (iter 22) — release-readiness umbrella (includes
mcp-scan)score-harness (iter 114) — broader 0-100 scorecard; MCP safety is
one of its 5 dimensionsrepo-genome (iter 114) — pre-scaffold readiness; MCP risk is one of
its 7 sectionsまだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Diff two scaffolded harnesses (ADR-031). Reports manifest meta drift + host list + per-file fingerprint changes (added/removed/changed). Exits 0 IDENTICAL, 1 DRIFT, 2 missing manifest. Use --bundle for the ADR-031 schema-1 JSON envelope.
日本語の概要は準備中です。原文の説明を表示しています。
Scaffold your own focused AI agent harness — pick host (Claude Code, Codex, pi.dev, Hermes), template, agents, skills, and ship a npm-publishable harness with its own npx CLI. Use when a user asks to "create my own agent harness", "scaffold a harness", "make a custom Claude Code plugin like ruflo", or "build a vertical AI assistant for X".
日本語の概要は準備中です。原文の説明を表示しています。
Kernel-version skew check (ADR-027). Reports manifest surface + manifest kernel + installed kernel + verdict (match/patch-diff/minor-diff/major-diff). Exits 1 on minor/major skew with a copy-pasteable `npm install @metaharness/kernel@X.Y.Z` next step. Exits 2 if no .harness/manifest.json at path.
日本語の概要は準備中です。原文の説明を表示しています。
Evolve this harness with Darwin Mode — frozen model, evolving harness (real, sandboxed, safety-gated).
日本語の概要は準備中です。原文の説明を表示しています。
Scaffold a ready-made AI agent harness in one command from the 19 published @metaharness/* example packages — 9 host integrations (Claude Code, Codex, Hermes, pi.dev, OpenClaw, RVM, Copilot, OpenCode, GitHub Actions) + 10 vertical pods (devops, research, trading, support, legal, coding, education, sales, gaming, repo-maintainer).
日本語の概要は準備中です。原文の説明を表示しています。
GCP Secret Manager integration: validate setup, fetch values, or confirm an NPM_TOKEN is non-revoked via `npm whoami`. Used for publish-time token rotation without long-lived keys in CI.
日本語の概要は準備中です。原文の説明を表示しています。