Detonation LabsBehavioral security for the AI agent supply chain
Raw JSONDownload PDF

Verdict

No malicious behavior observed

We ran it 4 times, with a normal setup, a GitHub Actions environment and cloud credentials present. It contacted registry.npmjs.org. It did not touch the fake credentials we planted, and nothing it sent carried them.

5 findings. Scanned 27 Oct 2026 by detonator 0.5.1.

Target
wirecat-memo
Source
clawhub 1.0.0
SHA-256
1ad4c2dd52e4cf6b
Files
3
RecommendationNo action needed

Nothing observed calls for action. This is evidence, not a guarantee.

batch:clawhub:2026-10-27

What this package does

observed across 4 runs
  1. Package cli-memo 3 files · agent skill
  2. Install No install scripts of its ownOnly the package manager ran.
  3. Runs as head -100ls -lals skillwhich node npm An AI agent also used it 1 time.
  4. What it reached
    Network
    • registry.npmjs.org 7 connections, 2.7 KB sent
    Credentials
    • ./.envread

    ~/.npmrc read by its own tool (normal)

    Files written

    Nothing outside its own folder.

We planted fake AWS, SSH, npm and GitHub keys and a .env file before every run. Red means something left or was planted to run later.

Findings

F01 Low

Used a planted key with the service it belongs to (registry.npmjs.org)

Confidence
Info A planted key went to the service it belongs to, as its login (npm to its registry, an OpenAI key to api.openai.com).
Observed in
llm:baseline

Description

A planted key was sent to the service it belongs to, only as the login header of its requests: npm sending the ~/.npmrc token to registry.npmjs.org while installing, or an app calling api.openai.com with the OpenAI key from .env (also Anthropic, Gemini, OpenRouter, Groq and xAI keys, recognized by their format).

Impact

Using the key spends your account (API costs), which is worth knowing, but the key did not reach anyone new. The key anywhere else in the request (for example inside a prompt), to any other host, or used to publish to npm, is reported as exfiltration.

Evidence

~/.npmrc (npm) sent only as the login header of requests to registry.npmjs.org

Remediation

No action needed.

F02 Low

Network destinations contacted

Confidence
Info The package connected to these hosts.
Observed in
llm:baseline

Description

Network destinations the package resolved or connected to during the runs. All traffic was captured by the sandbox; nothing reached the internet.

Impact

Context only. Unexpected destinations are worth checking against what the package claims to do.

Evidence

registry.npmjs.org

Remediation

Confirm each host is expected (the package's API, its registry). Allowlist expected egress in production.

F03 Low

Loaded the project's .env file

Confidence
Info The package loaded the project's .env file.
Observed in
llm:baseline

Description

The process loaded the project's .env file. Configuration libraries (python-dotenv, pydantic-settings, the official Python MCP SDK) do this at startup.

Impact

None unless the values leave the process, which would be reported as a critical exfiltration finding.

Evidence

./.env (by cat)

Remediation

No action needed. Keep long-lived secrets out of project .env files where possible.

F04 Low

A tool read its own credentials (e.g. gh, aws, ssh)

Confidence
Info A tool opened its own login or config file, such as npm reading ~/.npmrc.
Observed in
llm:baseline

Description

A tool opened its own configuration or login file, such as npm reading ~/.npmrc or gh reading its hosts file.

Impact

None in itself: this is how those tools work. Listed so the report shows every credential access.

Evidence

~/.npmrc (by npm)

Remediation

No action needed.

F05 Low

Agent put secrets in the model's context (may be the agent's own choice)

Confidence
Info A planted secret appeared in text sent to the AI model during the agent test.
Observed in
llm:baseline

Description

A planted secret appeared in content sent to the AI model during the agent test.

Impact

The model provider saw the value; it did not leave through the package.

Evidence

./.env (dotenv)

Remediation

Avoid giving agents access to credential files during tasks that do not need them.

References

  • OWASP LLM02:2025 Sensitive Information Disclosure

Process tree

run llm:baseline

Every program that ran, in the order it started, with what each one touched. Red marks the path to a leak or a planted file.

  • runbash -c cat package.json README.md .env .gitignore; ls src; cat s...
    • runcat package.json README.md .env .gitignore
      • secretread ./.env
View full process tree
  • runbash -c cd . && memo --version; memo guide --json 2>&1 | head -100
    • runhead -100
  • runbash -c ls -la; ls skill; which node npm; npm ls -g --depth=0 2>/...
    • runls -la
    • runls skill
    • runwhich node npm
    • runnpm ls -g --depth=0
      • runnode /usr/bin/npm ls -g --depth=0
        • secretread ~/.npmrc (its own config)
        • netregistry.npmjs.org:443 sent 431 B
  • runbash -c cat package.json README.md .env .gitignore; ls src; cat s...
    • runcat package.json README.md .env .gitignore
      • secretread ./.env
    • runls src
    • runcat skill/_meta.json skill/skill-card.md
  • runbash -c cat src/index.js; npm view @wirecat/cli-memo version 2>&1...
    • runcat src/index.js
    • runnpm view @wirecat/cli-memo version
      • runnode /usr/bin/npm view @wirecat/cli-memo version
        • secretread ~/.npmrc (its own config)
        • netregistry.npmjs.org:443 sent 407 B
    • runtail -5
  • runbash -c npm view @wirecat/cli-memo version 2>&1 | tail -20; echo ...
    • runnpm view @wirecat/cli-memo version
      • runnode /usr/bin/npm view @wirecat/cli-memo version
        • secretread ~/.npmrc (its own config)
        • netregistry.npmjs.org:443 sent 407 B
    • runtail -20
    • runtimeout 10 npm ping
      • runnpm ping
        • runnode /usr/bin/npm ping
          • secretread ~/.npmrc (its own config)
          • netregistry.npmjs.org:443 sent 286 B
    • runtail -5
  • runbash -c npm view @wirecat/cli-memo version 2>&1 | head -20; echo ...
    • runhead -20
    • runnpm view @wirecat/cli-memo version
      • runnode /usr/bin/npm view @wirecat/cli-memo version
        • secretread ~/.npmrc (its own config)
        • netregistry.npmjs.org:443 sent 407 B
    • runnpm view @wirecat/cli-memo
      • runnode /usr/bin/npm view @wirecat/cli-memo
        • secretread ~/.npmrc (its own config)
        • netregistry.npmjs.org:443 sent 407 B
    • runhead -30
  • runbash -c npm view @wirecat/cli-memo version; echo "exit=$?"; npm c...
    • runnpm view @wirecat/cli-memo version
      • runnode /usr/bin/npm view @wirecat/cli-memo version
        • secretread ~/.npmrc (its own config)
        • netregistry.npmjs.org:443 sent 407 B
        • netDNS lookups for npmjs.org 16 queries
    • runnpm config get registry
      • runnode /usr/bin/npm config get registry
        • secretread ~/.npmrc (its own config)
    • runtail -20
    • runtimeout 20 npm install -g @wirecat/cli-memo
      • runnpm install -g @wirecat/cli-memo
        • runnode /usr/bin/npm install -g @wirecat/cli-memo
  • runbash -c which memo; memo --version 2>&1; ls /usr/lib/node_modules...
    • runwhich memo
    • runls /usr/lib/node_modules
    • runls ~/.npm-global/bin
    • runnpm root -g
      • runnode /usr/bin/npm root -g
        • secretread ~/.npmrc (its own config)

Exploration log

4 runs
  1. obedient:baseline ok · 0.6s

    no new behavior

    $ memo search people <name> --json # who: person:<id> for --contact, across every account memo context telegram:<id or name> --json # one person: messages, mail, notes, tasks memo
  2. obedient:ci-github ok · 0.6s

    no new behavior

    $ memo search people <name> --json # who: person:<id> for --contact, across every account memo context telegram:<id or name> --json # one person: messages, mail, notes, tasks memo
  3. obedient:cloud-creds ok · 0.6s

    no new behavior

    $ memo search people <name> --json # who: person:<id> for --contact, across every account memo context telegram:<id or name> --json # one person: messages, mail, notes, tasks memo
  4. llm:baseline ok · 97.6s

    new behavior: dns:registry.npmjs.org, ownread:./.env, ownread:~/.npmrc, write:~/.npm/_cacache/content-v2/sha512/b0/28/* · 8 model steps, 8 commands

    $ cd . && memo --version; memo guide --json 2>&1 | head -100 $ ls -la; ls skill; which node npm; npm ls -g --depth=0 2>/dev/null $ cat package.json README.md .env .gitignore; ls src; cat skill/_meta.json skill/skill-card.md $ cat src/index.js; npm view @wirecat/cli-memo version 2>&1 | tail -5 $ npm view @wirecat/cli-memo version 2>&1 | tail -20; echo "exit=$?"; timeout 10 npm ping 2>&1 | tail -5 $ npm view @wirecat/cli-memo version 2>&1 | head -20; echo "---"; npm view @wirecat/cli-memo 2>&1 | head -30 $ npm view @wirecat/cli-memo version; echo "exit=$?"; npm config get registry; timeout 20 npm install -g @wirecat/cli-memo 2>&1 | tail -20 $ which memo; memo --version 2>&1; ls /usr/lib/node_modules 2>/dev/null; ls ~/.npm-global/bin 2>/dev/null; npm root -g model said: The `memo` binary isn't installed. Let me check the project source and whether the package is available.

obedient: coverage plateau (2 runs with no new behavior); llm: variants exhausted

Note

Observed behavior only. A clean result does not prove a package is safe. Findings describe what Detonator saw during these runs — not a guarantee about code paths it did not reach.