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 mcp.quantplay.tech. 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
quantplay
Source
clawhub 1.0.1
SHA-256
1cf5f19a79ef819b
Files
3
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 quantplay 3 files · agent skill
  2. Install No install stepFiles were copied in as they are.
  3. Runs as ls -la ./skillcat ./skill/SKILL.mdhead -50cat ./skill/_meta.json An AI agent also used it 1 time.
  4. What it reached
    Network
    • mcp.quantplay.tech 6 connections, 1.4 KB 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

mcp.quantplay.tech

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 README.md package.json .gitignore; echo ---; cat .env...
    • runcat .env
      • secretread ./.env
View full process tree
  • runbash -c ls -la ./skill && cat ./skill/SKILL.md | head -50
    • runls -la ./skill
    • runcat ./skill/SKILL.md
    • runhead -50
  • runbash -c cat ./skill/_meta.json; echo ---; cat ./skill/skill-card....
    • runcat ./skill/_meta.json
    • runcat ./skill/skill-card.md
    • runls -la
    • runenv
    • rungrep -i -E quantplay|mcp
    • runls ~/.config
    • runcat ~/.claude.json
    • runhead -50
  • runbash -c cat README.md package.json .gitignore; echo ---; cat .env...
    • runcat .env
      • secretread ./.env
    • runcat README.md package.json .gitignore
    • runsed s/=.*/=<redacted>/
    • runls -la src
    • rungit log --oneline
    • runhead
  • runbash -c cat src/index.js; echo ---; which curl node python3; echo...
    • runcat src/index.js
    • runwhich curl node python3
    • runcurl -s -m 15 -o /dev/null
      • netmcp.quantplay.tech:443 sent 93 B
  • runbash -c cd /tmp && curl -s -m 20 -D - -X POST https://mcp.quantpl...
    • runcurl -s -m 20 -D -
      • netmcp.quantplay.tech:443 sent 329 B
    • runhead -40
  • runbash -c cd /tmp && curl -s -m 20 -X POST https://mcp.quantplay.te...
    • runhead -c 3000
    • runcurl -s -m 20 -X POST
      • netmcp.quantplay.tech:443 sent 236 B
    • runcurl -s -m 20 -X POST
      • netmcp.quantplay.tech:443 sent 240 B
    • runhead -c 2000
  • runbash -c cd /tmp && curl -s -m 20 -X POST https://mcp.quantplay.te...
    • runhead -c 1500 ×2
    • runcurl -s -m 20 -X POST
      • netmcp.quantplay.tech:443 sent 254 B
    • runcurl -s -m 20 -X POST
      • netmcp.quantplay.tech:443 sent 240 B
      • netDNS lookups for quantplay.tech 12 queries

Exploration log

1 run
  1. llm:baseline ok · 21.9s

    new behavior: dns:mcp.quantplay.tech, exec:curl, ownread:./.env · 8 model steps, 7 commands

    $ ls -la ./skill && cat ./skill/SKILL.md | head -50 $ cat ./skill/_meta.json; echo ---; cat ./skill/skill-card.md; echo ---; ls -la; echo ---; env | grep -i -E 'quantplay|mcp' ; echo ---; ls ~/.config 2>/dev/null; cat ~/.claude.json 2>/dev/null | head -5 $ cat README.md package.json .gitignore; echo ---; cat .env | sed 's/=.*/=<redacted>/'; echo ---; ls -la src; echo ---; git log --oneline | head $ cat src/index.js; echo ---; which curl node python3; echo ---; curl -s -m 15 -o /dev/null -w '%{http_code}\n' https://mcp.quantplay.tech/mcp/readonly $ cd /tmp && curl -s -m 20 -D - -X POST https://mcp.quantplay.tech/mcp/readonly -H 'Content-Type: application/json' -H 'Accept: application/json, text/event-stream' -d '{"jsonrpc":"2.0","id":1,"method": $ cd /tmp && curl -s -m 20 -X POST https://mcp.quantplay.tech/mcp/readonly -H 'Content-Type: application/json' -H 'Accept: application/json, text/event-stream' -d '{"jsonrpc":"2.0","id":2,"method":"tool $ cd /tmp && curl -s -m 20 -X POST https://mcp.quantplay.tech/mcp/readonly -H 'Content-Type: application/json' -d '{"jsonrpc":"2.0","id":4,"method":"tools/call","params":{"name":"get_holdings","argument model said: I couldn't complete the task, and I want to be straight with you about why. **What I tried** - Read the skill at `./skill/SKILL.md`. It's a read-only connector to the Quantplay MCP server (`https://mcp.quantplay.tech/mcp/readonly`) that reads Indian broker accounts (positions, holdings, orders, margins, P&L, option chains). - The skill's tools are supposed to be exposed to me as MCP tools. **They

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.