agent-comms-mcp@0.9.6
VERIFIED· not startedPublisher claims: "MCP service for permissioned, structured agent-to-agent communications" Observed in sandbox: no network, touches nothing outside the project, writes no memory.
Coverage
Not started: dependency install failed; the server itself was not observed.- Tested
- the dependency install step · every file read, file written and host contacted by what ran
- Not tested
- server startup · the tool list · tool behavior
What this means for the label: Verified here covers what ran. The server's tools were not exercised with a real credential, so nothing about their behavior is asserted; the credential summary below says what to do with that.
Before you give it your token
- It will ask for
DATABASE_URLMCP_TOKEN_STORAGE_PATHConfiguration, not credentials: _TEST_FORCE_INVALID_SCRATCH_INDEX, _TEST_SIZE_GUARD_THRESHOLD, BASE_URL, MCP_HOST, MCP_PORT- Where it talks to
- no vendor host named in the codeAlso arcadia-dev.example, example.org — not the publisher's domain; check what for before connecting.
- What we saw when it ran
- no credential files read · no network
- What we could not test
- server startup · the tool list · tool behavior
Connect it with a scoped credential, under the guard, and read the third-party hosts first. The code names arcadia-dev.example, example.org as well as its own vendor. The guard injects DATABASE_URL, MCP_TOKEN_STORAGE_PATH only into requests to the host you bind (and its subdomains) and refuses the rest.
Never paste a credential into a chat, a ticket or a report, including ours. How the guard holds it.
What changed since 0.9.3
vet-skills diff agent-comms-mcp 0.9.3 0.9.6This version reaches for something 0.9.3 did not: 2 new hosts (arcadia-dev.example, example.org). Compared from the two recorded runs, not from source.
- New hosts
- arcadia-dev.example, example.org
0.9.3 was not run in the sandbox (install failed); the comparison covers what static analysis recorded only.
The guard can hold an update like this until someone approves it (runtime.on_change: hold). A change that only shows with a real credential, or on a path the run did not reach, is not in a trace and so not in this comparison.
Is [email protected] safe to install?
KillerSkills ran it in the KillerSkills honeypot sandbox (2 scenarios; not started: dependency install failed; the server itself was not observed) on 2026-11-22 and scored it 82/100 (Verified). In what ran, nothing we could detect went wrong: no undeclared network egress, no credential reads, no persistent instruction writes. What it does with a real credential was not observed; the static findings and the publisher carry the rest of the score. Summary: no network, touches nothing outside the project, writes no memory.
Behavior trace
Honeypot run · 1 s · 2 scenarios · 0 deviationsStatic findings
"See .env.example for the full list of required configuration."Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.res_up = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.res_up2 = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.res_down = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.res_down = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.res_fail = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.res_up = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.res_up1 = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.res = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out."http://10.0.0.1:1055",Fix: Replace encoded or assembled strings with plain literals and remove undeclared network calls.result = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.res_down2 = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.res_up = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.res_rerun_ok = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.request = client.build_request("GET", "https://10.0.0.1/actions/proposals/apply")Fix: Replace encoded or assembled strings with plain literals and remove undeclared network calls.assert decide_resp.json()["sender_agent_id"] == str(agent.id)Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.res_up = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.res = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.# The kill switch guard sets decision_note on hold before committing under the row lock.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.…ms:write` (additionally requires `comms:admin` or an interactive/Okta caller) | Admin override of an existing agent's `is_shared` value, since `comms_register` freezes it aga…Fix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out."""TECH-5736 suggestion: suspension is meant to be a real killFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.res_down = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out."https://user:[email protected]", # userinfoFix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.# Returned proposal reflects human decision, NOT kill switch downgradeFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.res_down1 = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.assert get_resp.json()["sender_agent_id"] == str(agent.id)Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.var blank by accident (e.g. a blank line in a ``.env`` file)."""Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.subprocess.run(["docker", "rm", "-f", container_id], capture_output=True, check=False)Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.result = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.res_ok = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.# Log configured proposal auto-approval kill switch and daily caps once at startup (TECH-8080 PR13).Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.meant to be a real kill switch, but this function -- called on everyFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.``tests/test_proposal_service.py``. Covers the kill switch, targetFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.res = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out."file:///etc/passwd",Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.proc = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.proc = subprocess.run(cmd, capture_output=True, text=True, check=True)Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.port_proc = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(["docker", "rm", "-f", pg.container_id], capture_output=True, check=False)Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.a ``.env`` file)."""Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.res_failed = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out."http://192.168.1.1:1055",Fix: Replace encoded or assembled strings with plain literals and remove undeclared network calls.Cost to the agent
Tokens estimated at four characters each. The tool list is re-read by the agent on every turn, so it is the number to compare between servers that do the same job: ranked across the index.
Known vulnerabilities
Not checked against OSV yet; the weekly check has not reached this version.