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

Verdict

No malicious behavior observed

We ran it 8 times, with a normal setup, WATCH_REPORT_DEPENDENCIES set, a GitHub Actions environment, the clock moved a year ahead and AGENT_TIMEOUT_MS and BASELINE and CURL_CA_BUNDLE and DENO_CERT… set. It made no network connections. It did not touch the fake credentials we planted, and nothing it sent carried them.

0 findings. Scanned 26 Oct 2026 by detonator 0.5.1.

Target
cgartlab-quill
Source
clawhub 0.11.0
SHA-256
9ef4ec651c16963c
Files
53
RecommendationNo action needed

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

batch:clawhub:2026-10-26

What this package does

observed across 8 runs
  1. Package quill 53 files · agent skill
  2. Install No install stepFiles were copied in as they are.
  3. Runs as node <skill>/evals/check-workspace.mjsnode <skill>/evals/run.mjs/usr/bin/node <skill>/evals/check-workspace.mjs --workspace <skill>/results/workspaces/eva/usr/bin/node <skill>/evals/check-workspace.mjs --eval 1 --workspace <skill>/results/works
  4. What it reached
    Network

    No connections.

    Credentials

    None of the planted fake keys were touched.

    Files written
    • ./.quill-snapshot.jsonwritten
    • ./true/coverage-13-1793042783613-0.jsonwritten
    • ./true/coverage-13-1793042786078-0.jsonwritten

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

No findings. Detonator observed no malicious or suspicious behavior in the runs below.

Process tree

run scripts: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.

Show process tree
  • runnode <skill>/evals/check-workspace.mjs
    • filewrote ./.quill-snapshot.json

Exploration log

8 runs
  1. scripts:baseline ok · 1.2s

    new behavior: write:./.quill-snapshot.json

  2. scripts:guided:WATCH_REPORT_DEPENDENCIES ok · 1.7s

    no new behavior

  3. scripts:ci-github ok · 3.1s

    no new behavior

  4. scripts:baseline timeout · 20.3s

    no new behavior

  5. scripts:time-travel timeout · 20.3s

    no new behavior

  6. scripts:guided:AGENT_TIMEOUT_MS,BASELINE,CURL_CA_BUNDLE,DENO_CERT… ok · 1.2s

    new behavior: write:./true/coverage-13-*-0.json

  7. scripts:guided:AGENT_TIMEOUT_MS,BASELINE,CURL_CA_BUNDLE,DENO_CERT… ok · 0.8s

    no new behavior

  8. scripts:ci-github timeout · 20.3s

    no new behavior

scripts/evals/check-workspace.mjs: coverage plateau (2 runs with no new behavior); scripts/evals/run.mjs: run budget spent; llm: run budget spent

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.