Proof of Commitment

Stars lie. Behavioral signals don't.

An MCP server and web tool that scores npm packages, PyPI packages, Rust crates, Go modules, and GitHub repos on behavioral commitment — signals that are harder to fake than stars, READMEs, or download counts.

$ npx proof-of-commitment axios zod chalk lodash minimatch
Scoring 5 npm packages... done in 3.0s

Package      Risk          Score   Publishers   Downloads      Age    Provenance
chalk        🔴 CRITICAL   72      1            432.9M/wk      14.6y  —
minimatch    🔴 CRITICAL   78      1            634.1M/wk      14.9y  —
lodash       🔴 CRITICAL   80      1            158.9M/wk      14.1y  —
zod          🔴 CRITICAL   83      1            161.2M/wk      6.3y   🔐 verified
axios        🔴 CRITICAL   88      1            115.7M/wk      11.8y  🔐 verified
                ⚠ COMPROMISED — axios token theft (2026-03-30)

⚠  5 CRITICAL packages found.
   CRITICAL = sole npm publisher + >10M weekly downloads (publish-access concentration risk)

npm audit flags none of these. They're not vulnerabilities — they're attack-surface concentration. One stolen npm token, one phished maintainer, and a single push reaches the whole ecosystem (axios, March 30 2026 — happened).

The supply chain problem

26 of the 91 npm packages with >10M weekly downloads have a single npm publisher. Together they account for over 3 billion downloads per week. npm audit doesn't surface this. Stars don't either.

Four packages in a typical Node.js project are CRITICAL right now:

• chalk — 432M downloads/week, 1 npm publisher
• zod — 185M downloads/week, 1 npm publisher (30+ GitHub contributors)
• lodash — 156M downloads/week, 1 npm publisher
• axios — 113M downloads/week, 1 npm publisher (attacked March 30, 2026)

They won't appear in your package.json either — but these are in almost every project:

• minimatch — 625M downloads/week, 1 npm publisher
• glob — 366M downloads/week, 1 npm publisher
• cross-spawn — 215M downloads/week, 1 npm publisher

Behavioral signals surface this. Stars and READMEs don't.

Quick install (MCP)

No login required. Add to any MCP-compatible AI tool and start querying supply chain risk.

Claude Desktop

Open ~/Library/Application Support/Claude/claude_desktop_config.json on macOS (config file reference) or %APPDATA%\Claude\claude_desktop_config.json on Windows, then add:

{
  "mcpServers": {
    "commit": {
      "type": "streamable-http",
      "url": "https://poc-backend.amdal-dev.workers.dev/mcp"
    }
  }
}

Restart Claude Desktop. A tool icon appears in the chat input — ask it to audit your package.json.

Cursor

Open ~/.cursor/mcp.json (Cursor MCP docs) and add:

{
  "mcpServers": {
    "commit": {
      "type": "streamable-http",
      "url": "https://poc-backend.amdal-dev.workers.dev/mcp"
    }
  }
}

Smithery (once indexed)

npx -y @smithery/cli install proof-of-commitment --client claude

---

Try it now

Terminal (zero install):

# New in v1.8.0: zero-arg auto-detect — cd into any project, run once:
npx proof-of-commitment
# Picks the highest-coverage manifest in cwd (package-lock.json > yarn.lock >
# pnpm-lock.yaml > pnpm-workspace.yaml > package.json; requirements.txt;
# Cargo.toml; go.sum > go.mod). When multiple ecosystems are present, the
# file with the most recent mtime wins.

# Explicit package list still works:
npx proof-of-commitment axios zod chalk

# Or point at a specific file:
npx proof-of-commitment --file package.json
npx proof-of-commitment --file package-lock.json   # npm (transitive)
npx proof-of-commitment --file yarn.lock           # yarn
npx proof-of-commitment --file pnpm-lock.yaml      # pnpm
npx proof-of-commitment --file pnpm-workspace.yaml # pnpm monorepo
npx proof-of-commitment --pypi litellm langchain requests
npx proof-of-commitment --cargo serde tokio reqwest
npx proof-of-commitment --golang github.com/gin-gonic/gin golang.org/x/net
npx proof-of-commitment --file go.mod
npx proof-of-commitment --file go.sum              # full transitive Go set

# JSON output for downstream tools:
npx proof-of-commitment --file package-lock.json --json | jq '.criticalCount'

CI integration (v1.8.0+)

--fail-on=<level> turns the CLI into a one-line CI gate. No GitHub Action required.

# .github/workflows/supply-chain.yml
name: Supply Chain
on: [pull_request]
jobs:
  audit:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-node@v4
        with: { node-version: '20' }
      - run: npx -y proof-of-commitment --fail-on=critical

Levels:

--fail-on	Exit 1 when…
critical	any package is flagged CRITICAL (publish-access concentration)
risky	any package is CRITICAL or HIGH (score < 40)
none	never — report only

Defaults: critical in CI (when CI=true is set, which every major CI runner does) and for --json output. Interactive (TTY, non-CI) keeps the v1.7 default of exit 0 — running locally won't break your shell habits.

The dedicated piiiico/commit-action@v1 is still the right choice when you want PR comments and step summaries; --fail-on is for minimal pipelines that just need a yes/no answer.

SARIF output for GitHub Code Scanning (v1.26.0+)

--sarif outputs SARIF 2.1.0 — the standard format for static analysis results. Upload it to GitHub Code Scanning and Commit findings appear in the Security tab alongside CodeQL and Snyk.

# .github/workflows/supply-chain.yml
name: Supply Chain
on: [pull_request]
jobs:
  audit:
    runs-on: ubuntu-latest
    permissions:
      security-events: write
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-node@v4
        with: { node-version: '20' }
      - run: npx -y proof-of-commitment --file package-lock.json --sarif --fail-on=none > results.sarif
      - uses: github/codeql-action/upload-sarif@v3
        if: always()
        with:
          sarif_file: results.sarif
          category: commit-supply-chain

CRITICAL and HIGH packages show as alerts in the repo's Security tab. Compromised packages (in the Commit incident registry) get a separate alert. --fail-on still controls the exit code independently — use --fail-on=critical to also block the PR.

Web demo (no install): getcommit.dev/audit — paste your packages, see risk scores in seconds.

IDE Hooks (Cursor + Claude Code + Windsurf)

poc hook installs a supply chain gate for Cursor (beforeShellExecution), Claude Code (PreToolUse), and Windsurf (pre_run_command) in one command. The same hook script intercepts package installs from any agent, auto-detects which client called it, and blocks CRITICAL packages before they run.

# Install for the current project (writes .cursor/hooks.json + .claude/settings.json + .windsurf/hooks.json):
poc hook

# Or protect every project for your user:
poc hook --global

# Narrow to one client:
poc hook --cursor          # only .cursor/hooks.json
poc hook --claude-code     # only .claude/settings.json
poc hook --windsurf        # only .windsurf/hooks.json

# Remove (cleans all three):
poc hook --uninstall

The hook writes .cursor/hooks.json, .claude/settings.json, and .windsurf/hooks.json (project) or the equivalents under ~/ (with --global). When Cursor, Claude Code, or Windsurf runs npm install axios, pip install litellm, cargo add serde, or go get github.com/gin-gonic/gin, the hook calls the Commit API and either blocks, warns, or allows — in under 500ms.

What gets intercepted:

Package manager	Example command
npm / npx	npm install <pkg>, npm add <pkg>
pnpm	pnpm add <pkg>
yarn	yarn add <pkg>
pip / pip3 / uv	pip install <pkg>
cargo	cargo add <pkg>, cargo install <pkg>
go	go get <module>, go install <module>

Why this matters: Supply chain attacks now happen in minutes. The Shai-Hulud worm (May 2026) compromised 637 packages in 39 minutes and specifically targeted AI coding assistants — planting persistence hooks in .claude/settings.json and .vscode/tasks.json. When your AI assistant installs a dependency, it bypasses the human review that used to be the last line of defense. poc hook puts a gate back in — same gate, whether Cursor, Claude Code, or Windsurf is driving.

Default behavior: CRITICAL packages (sole npm publisher + >10M downloads/week — the exact LiteLLM/axios attack profile) are blocked. HIGH packages trigger an "ask user" prompt (Cursor/Claude Code) or are blocked with a message (Windsurf). Set COMMIT_HOOK_SEVERITY_BLOCK=HIGH to block both.

With an API key: poc login sk_commit_… before running poc hook — the key is embedded in the hook config and lifts the rate limit.

---

Get notified before the next attack

The CLI tells you what's risky today. A free API key unlocks monitoring — score recomputation across the packages you depend on, with alerts when one degrades (publisher drops, release stalls, score falls ≥10 points).

• Open (free): Watch 3 packages · weekly digest every Monday
• Developer ($15/mo): Watch 15 packages · daily scans · instant email alerts

Get a free API key → (no card, 30 seconds · 200 audits/day included)

npm install -g proof-of-commitment   # then:
poc watch axios --email you@company.com  # free key + monitoring in one step
poc watch chalk                          # add more packages (3 free)
poc init                                 # add CI gate to this repo

GitHub Action

Add supply chain auditing to any CI pipeline in 30 seconds — auto-detects packages from package.json or requirements.txt, posts results as a PR comment, writes to GitHub Step Summary, and optionally fails on CRITICAL packages.

Use the dedicated action at piiiico/commit-action:

# .github/workflows/supply-chain.yml
name: Supply Chain Audit
on:
  pull_request:
    paths: ['package.json', 'package-lock.json', 'bun.lock']

jobs:
  audit:
    runs-on: ubuntu-latest
    permissions:
      pull-requests: write
    steps:
      - uses: actions/checkout@v4
      - uses: piiiico/commit-action@v1
        with:
          fail-on-critical: true   # blocks merges on CRITICAL packages
          comment-on-pr: true      # posts results as a PR comment

When comment-on-pr: true (default), the action automatically posts the audit table as a comment on the pull request — and updates the same comment on re-run, so you don't get comment spam. Reviewers see the risk table without leaving the PR.

Inputs:

Input	Default	Description
packages	(auto)	Comma-separated package names (auto-detected from package.json/requirements.txt if not set)
packages-file	(auto)	Path to package.json or requirements.txt (default: auto-detect in workspace root)
fail-on-critical	true	Fail the workflow if CRITICAL packages are found
max-packages	20	Max packages to audit when auto-detecting
include-dev-dependencies	false	Include devDependencies from package.json
comment-on-pr	true	Post audit results as a PR comment (requires pull-requests: write permission)
api-key	(none)	Commit Pro API key — enables batch requests and 10K requests/month
api-url	(prod)	Override API endpoint (useful for self-hosting)

Outputs: has-critical, critical-count, audit-summary (markdown table, also written to Step Summary).

Free vs Pro: Without an API key, packages are audited one at a time (with delays to respect rate limits). With a Pro API key, all packages are audited in a single batch request — faster and with higher monthly limits.

Example PR comment / Step Summary output:

| Package | Risk        | Score | Publishers | Downloads/wk | Age   |
|---------|-------------|-------|------------|--------------|-------|
| chalk   | 🔴 CRITICAL | 75    | 1          | 380M         | 12.7y |
| zod     | 🔴 CRITICAL | 83    | 1          | 133M         | 6.1y  |
| axios   | 🔴 CRITICAL | 89    | 1          | 93M          | 11.6y |

README Badges

Add a Commit Trust badge to any npm package you maintain or depend on:

![Commit Trust](https://poc-backend.amdal-dev.workers.dev/badge/YOUR-PACKAGE)

Examples:

Package	Badge URL
chalk	
react	
express	
@babel/core	

Grades: 🟢 OK (75+) · 🟠 WARNING (40–74) · 🔴 CRITICAL (<40 or sole npm publisher with 10M+ weekly downloads)

Badges are cached 1 hour. No API key needed.

Also supports PyPI, Cargo, Go modules, and the full ecosystem-specific format:

![commit score](https://poc-backend.amdal-dev.workers.dev/api/badge/npm/YOUR-PACKAGE)
![commit score](https://poc-backend.amdal-dev.workers.dev/api/badge/pypi/YOUR-PACKAGE)
![commit score](https://poc-backend.amdal-dev.workers.dev/api/badge/cargo/YOUR-CRATE)
![commit score](https://poc-backend.amdal-dev.workers.dev/api/badge/golang/github.com/owner/repo)

REST API

No API key. No install.

curl https://poc-backend.amdal-dev.workers.dev/api/audit \
  -X POST \
  -H "Content-Type: application/json" \
  -d '{"packages": ["axios", "zod", "chalk", "lodash", "express"]}'

{
  "count": 5,
  "results": [
    {
      "name": "chalk",
      "ecosystem": "npm",
      "score": 75,
      "maintainers": 1,
      "weeklyDownloads": 398397580,
      "ageYears": 12.7,
      "trend": "stable",
      "riskFlags": ["CRITICAL"],
      "scorecardScore": 3.6,        // null if no GitHub repo
      "hasDangerousWorkflow": false  // null if no Scorecard data
    },
    ...
  ]
}

11 MCP tools

Tool	Description
audit_dependencies	Batch risk audit for up to 20 npm/PyPI/Cargo/Go packages
audit_github_repo	Fetch a repo's package.json/requirements.txt and audit every dep
audit_dependency_tree	Map an npm package's full dependency tree (incl. transitive CRITICAL deps)
lookup_npm_package	Single npm package behavioral profile
lookup_pypi_package	Single PyPI package behavioral profile
lookup_cargo_crate	Single Rust crate behavioral profile (crates.io)
lookup_go_module	Single Go module behavioral profile (proxy.golang.org + GitHub)
lookup_github_repo	GitHub repo commitment score (longevity, commit frequency, contributor depth)
lookup_business	Norwegian business register — operating years, employees, financials
lookup_business_by_org	Same, by org number
query_commitment	Browser extension behavioral data (unique verified visitors, repeat rate)

Anonymous: 15 requests/IP/UTC day across both /mcp and /api/audit. Free key (no card, 30s signup at https://getcommit.dev/get-started): 200/day. Higher tiers at https://getcommit.dev/pricing.

What the score measures

Each package is scored 0–100 across:

• Longevity — How long has the package existed? Abandoned packages get reactivated for attacks.
• Publisher depth — Single npm publisher + millions of weekly downloads = the attack surface LiteLLM exploited. (Publisher = person with npm publish access, distinct from GitHub contributors.)
• Release consistency — Regular releases signal active oversight. Long gaps = vulnerability accumulation.
• Download trend — Growing packages attract more scrutiny (and attacks). Stable = lower profile.
• OpenSSF Scorecard — Process security (code review enforcement, branch protection, CI/CD safety). Separate from behavioral signals. High Scorecard ≠ safe from credential theft attacks.

Both axios (8.1/10 Scorecard) and chalk (3.6/10 Scorecard) score CRITICAL on behavioral signals. They measure different attack surfaces — Scorecard catches process gaps, behavioral signals catch publisher concentration.

Risk flags:

• CRITICAL — single npm publisher + >10M weekly downloads (exact LiteLLM/axios attack profile)
• HIGH — package <1yr old + rapid adoption
• WARN — no release in 12+ months

Real data points

# packages you know about:
chalk       — score 75, 1 publisher, 432M/week  ⚑ CRITICAL
zod         — score 83, 1 publisher, 185M/week  ⚑ CRITICAL  (30+ GitHub contributors)
lodash      — score 81, 1 publisher, 156M/week  ⚑ CRITICAL
axios       — score 88, 1 publisher, 113M/week  ⚑ CRITICAL  (attacked Mar 30 2026)
express     — score 90, 5 publishers, 95M/week

# packages probably not in your package.json, definitely in your lock file:
minimatch   — score 78, 1 publisher, 625M/week  ⚑ CRITICAL
glob        — score 80, 1 publisher, 366M/week  ⚑ CRITICAL
cross-spawn — score 72, 1 publisher, 215M/week  ⚑ CRITICAL

# post-attack:
litellm     — score 74, 1 publisher            ⚑ CRITICAL  (supply chain attack Mar 2026)

# Rust crates (new in v1.3.0):
serde       — score 78, 1 owner,  13M/week  ⚑ CRITICAL  (dtolnay sole owner)
tokio       — score 89, 2 owners, 10M/week
reqwest     — score 85, 1 owner,   8M/week  ⚑ HIGH

Why behavioral signals

The LiteLLM attack (March 2026) and axios attack (March 30, 2026) followed the same pattern: stolen credentials → malicious package pushed → 97M+ machines exposed. Both packages scored CRITICAL by these metrics before the attacks.

Declarative signals (stars, README quality, CI badges) don't capture this risk. Behavioral commitment does.

Blog

• The LinkedIn Backdoor: Why npm audit Missed a 250-Line Attack — A fake recruiter clone attack hid malware in test files and ran it via npm lifecycle scripts. npm audit: silent. What behavioral signals would have flagged.
• Axios Attack Prediction — We flagged axios as CRITICAL (sole npm publisher, 113M downloads/week) before the March 30, 2026 token theft.

Stack

Layer	Technology
Backend	Cloudflare Workers + D1
MCP	Model Context Protocol SDK
Data	npm registry, PyPI, crates.io, proxy.golang.org, deps.dev, GitHub API, Brønnøysund (NO)
Landing	Astro + Cloudflare Pages

Roadmap

Planned, not promised. The project is early-stage — contributions welcome on any of these.

Feature	Status	Notes
Cargo (Rust) registry support	✅ Live	MCP tool, REST API, badge endpoint — ecosystem: "cargo"
Go modules support	✅ Live	proxy.golang.org + deps.dev + GitHub-primary scoring — ecosystem: "golang"
Score breakdown visualization	Planned	Chart component for the 5 dimensions on getcommit.dev/audit
--json flag for CLI	✅ Live	npx proof-of-commitment --file package-lock.json --json | jq '.criticalCount'
pnpm workspace monorepo support	✅ Live	--file pnpm-workspace.yaml or auto-detected from pnpm-lock.yaml
Historical score tracking	Planned	Trend charts — was this package getting riskier over time?
Org-level dashboards	Planned	Aggregate risk view across all repos in a GitHub org

See open issues for things you can help with today.

The broader vision

Supply chain auditing is the first tool. The underlying primitive is a commitment graph — behavioral signals that replace content-based trust across any domain.

When content is free to fake (reviews, stars, READMEs), commitment becomes the signal. A publisher who has shipped 847 releases over 12 years is a different kind of commitment than one who published once in 2023.

The same logic applies to websites, businesses, and AI agents. Two card networks have independently named this gap: Mastercard Verifiable Intent §9.2 explicitly lists behavioral trust as "not covered." Visa TAP identifies agents without answering whether to trust them.

Proof of Commitment is the trust layer they're pointing at.

→ getcommit.dev

Run locally

bun install
bun run dev:backend     # local server with SQLite
bun run test:e2e        # E2E test with mock World ID

Deploy:

bun run deploy          # deploys to Cloudflare Workers

Releasing

Publish is triggered automatically when a tag v* is pushed, or manually via GitHub Actions workflow_dispatch.

Funnel smoke gate

Before npm publish runs, the CI workflow executes scripts/funnel-smoke.sh — a local-mock pre-publish check that exercises four key funnel paths:

Path	What it tests	Bug class caught
A	CLI audit with COMMIT_API_KEY set → 200 + results	v1.20.0: missing Authorization header → 0 paid conversions
B	CLI audit anonymous, 429 → message + instant_key_url	429 handling / CTA surfacing
C	cursor-hook (Cursor stdin) 429 → permission: ask + signup URL	v1.21.0: silent allow on 429 → security gap + 0 conversions
D	cursor-hook (Claude Code PreToolUse stdin) 429 → hookSpecificOutput.permissionDecision: ask + claude-code-hook-429 attribution	v1.22.0: wrong-shape reply when Claude Code drives → silent allow / mis-attributed conversion

Any path failure blocks the release. The gate runs a local Python mock server so it's deterministic in CI and doesn't depend on production rate-limit state.

Optional CI secret: Set COMMIT_TEST_API_KEY in GitHub repo secrets to use a real API key for Path A. Falls back to a mock key that the local server accepts unconditionally.

Run locally:

bash scripts/funnel-smoke.sh