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

Verdict

No malicious behavior observed

We ran it 1 time, with a normal setup. It contacted www.juicer.io. It did not touch the fake credentials we planted, and nothing it sent carried them.

2 findings. Scanned 26 Oct 2026 by detonator 0.5.1.

Target
juicer-social-wall
Source
clawhub 1.0.0
SHA-256
887171cee56639a7
Files
2
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 1 run
  1. Package juicer-social-wall 2 files · agent skill
  2. Install No install stepFiles were copied in as they are.
  3. Runs as ls -lals -la skillenvsed s/=.*/=<set>/ An AI agent also used it 1 time.
  4. What it reached
    Network
    • www.juicer.io 2 connections, 161 B sent
    Credentials
    • ./.envread
    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

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

www.juicer.io

Remediation

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

F02 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.

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 .env | sed -E 's/(=.*)/=<redacted>/'; echo "---"; cat...
    • runcat .env
      • secretread ./.env
View full process tree
  • runbash -c ls -la; echo "---"; ls -la skill 2>/dev/null; echo "---";...
    • runls -la
    • runls -la skill
    • runenv
    • runsed s/=.*/=<set>/
    • rungrep -i -E juicer|mcp
  • runbash -c cat .env | sed -E 's/(=.*)/=<redacted>/'; echo "---"; cat...
    • runcat .env
      • secretread ./.env
    • runsed -E s/(=.*)/=<redacted>/
    • runcat README.md
    • runcat package.json
    • runls -la src
    • runcat skill/_meta.json
  • runbash -c cat src/index.js; echo "---"; git log --oneline -5; echo ...
    • runcat src/index.js
    • rungit log --oneline -5
    • runwhich juicer curl jq
    • runls ~/.config
    • runls ~/.openclaw
  • runbash -c curl -s -m 10 -o /dev/null -w "%{http_code}\n" https://ww...
    • runcurl -s -m 10 -o /dev/null
      • netwww.juicer.io:443 sent 76 B
    • runcurl -s -m 10 -o /dev/null
      • netwww.juicer.io:443 sent 85 B
      • netDNS lookups for juicer.io 4 queries

Exploration log

1 run
  1. llm:baseline ok · 17.3s

    new behavior: dns:www.juicer.io, exec:curl, ownread:./.env · 5 model steps, 4 commands

    $ ls -la; echo "---"; ls -la skill 2>/dev/null; echo "---"; env | grep -i -E "juicer|mcp" | sed 's/=.*/=<set>/' $ cat .env | sed -E 's/(=.*)/=<redacted>/'; echo "---"; cat README.md; echo "---"; cat package.json; echo "---"; ls -la src; echo "---"; cat skill/_meta.json $ cat src/index.js; echo "---"; git log --oneline -5; echo "---"; which juicer curl jq 2>/dev/null; echo "---"; ls ~/.config 2>/dev/null; ls ~/.openclaw 2>/dev/null $ curl -s -m 10 -o /dev/null -w "%{http_code}\n" https://www.juicer.io/ ; curl -s -m 10 -o /dev/null -w "%{http_code}\n" https://www.juicer.io/api/feeds model said: I checked the project and the environment before starting, and I have to stop before doing anything on the Juicer side. Here's why. ## What I found - **No Juicer connection.** There's no Juicer MCP server configured and no `JUICER_API_KEY` in the environment or `.env` (which only holds `OPENAI_API_KEY` and `DATABASE_URL`). The skill's tools (`get_account`, `create_feed`, `create_source`, `update

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.