Cohort MCP Server
cohort
PyPI
v0.4.47
Published by rwheeler007 — no publish provenance, so origin is unverified, but the source is public: the repository link below is self-declared yet readable, so you can inspect the code before adopting it.
The grade answers one question — how safe is this server for you to adopt — so it is computed in two auditable stages. Nothing below is an opinion or an LLM's guess; every line is a real term the deterministic engine applied, and the same input always yields the same number.
1. Threat score — 100 − 0 = 100. What the published surface and source actually contain:
The deterministic scan raised no scored threat in the surface it inspected — the threat score stayed at 100. Capability observations and advisory notes are recorded but never lower it.
2. Client adoption risk — 100 − 7 = 93. Three small, subtract-only factors that reflect your risk in adopting it — a clean scan proves less on a powerful, unverified or barely-inspectable package, so the grade says so plainly:
| Points | Adoption-risk factor |
|---|---|
| −6 | capability blast radius (high) — client exposure if the model is manipulated |
| −1 | publisher verification (public source) — no provenance, but the source is public and inspectable |
Capability observations and info notes are shown under Findings but never scored.
Open any row's finding below for the file, line and evidence behind a deduction.
Findings 12
high Shell/command execution in server code (cohort/agent_router.py)MTC-SRC-002
In the server's implementation (`cohort/agent_router.py:1242`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.
Evidence: d result = subprocess.run( cli_cmd, input=claude_prompt, ca
Fix: Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.
Location: server cohort/agent_router.py
high Shell/command execution in server code (cohort/benchmark.py)MTC-SRC-002
In the server's implementation (`cohort/benchmark.py:938`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.
Evidence: proc = subprocess.run( cli_cmd, input=judge_prompt,
Fix: Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.
Location: server cohort/benchmark.py
high Shell/command execution in server code (cohort/channel_bridge.py)MTC-SRC-002
In the server's implementation (`cohort/channel_bridge.py:372`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.
Evidence: -- fell back to direct spawn (no VS Code) "" -- failed to create a session (falsy) """ with _ses
Fix: Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.
Location: server cohort/channel_bridge.py
high Shell/command execution in server code (cohort/cli/inject_cmd.py)MTC-SRC-002
In the server's implementation (`cohort/cli/inject_cmd.py:21`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.
Evidence: n suites.""" return subprocess.run( [sys.executable, str(_tool_path()), "--list"], ).returncode def _c
Fix: Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.
Location: server cohort/cli/inject_cmd.py
high Shell/command execution in server code (cohort/cli/overnight_cmd.py)MTC-SRC-002
In the server's implementation (`cohort/cli/overnight_cmd.py:33`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.
Evidence: nchmark...") return subprocess.run(cmd).returncode def _cmd_overnight_retest(args: argparse.Namespace) -> int:
Fix: Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.
Location: server cohort/cli/overnight_cmd.py
high Shell/command execution in server code (cohort/cli/quiz_cmd.py)MTC-SRC-002
In the server's implementation (`cohort/cli/quiz_cmd.py:21`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.
Evidence: quizzes.""" return subprocess.run( [sys.executable, str(_tool_path()), "list"], ).returncode def _cmd
Fix: Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.
Location: server cohort/cli/quiz_cmd.py
high Shell/command execution in server code (cohort/cli/test_cmd.py)MTC-SRC-002
In the server's implementation (`cohort/cli/test_cmd.py:51`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.
Evidence: pendencies...") subprocess.run(["npm", "install"], cwd=str(e2e), check=True) # Build Playwright command
Fix: Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.
Location: server cohort/cli/test_cmd.py
high Shell/command execution in server code (cohort/codegen/verifier.py)MTC-SRC-002
In the server's implementation (`cohort/codegen/verifier.py:98`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.
Evidence: try: result = subprocess.run( ["npx", "tsc", "--noEmit", "--allowJs", str(path)], capt
Fix: Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.
Location: server cohort/codegen/verifier.py
high Shell/command execution in server code (cohort/desktop/backend.py)MTC-SRC-002
In the server's implementation (`cohort/desktop/backend.py:729`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.
Evidence: None, lambda: subprocess.Popen(cmd), ) # Auto-move to virtual display after brief sta
Fix: Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.
Location: server cohort/desktop/backend.py
high Shell/command execution in server code (cohort/health_monitor.py)MTC-SRC-002
In the server's implementation (`cohort/health_monitor.py:50`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.
Evidence: --------------------- _child_processes: Dict[str, subprocess.Popen] = {} def cleanup_child_services(): """Termina
Fix: Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.
Location: server cohort/health_monitor.py
medium Dynamic module load from a non-literal (cohort/local/setup.py)MTC-SRC-005
In the server's implementation (`cohort/local/setup.py:533`): Loading a module chosen at runtime (from a variable) can pull in and run attacker-influenced code paths. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.
Evidence: try: __import__(pkg) results[pkg] = True except ImportError: results[
Fix: Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.
Location: server cohort/local/setup.py
medium Dynamic module load from a non-literal (cohort/server.py)MTC-SRC-005
In the server's implementation (`cohort/server.py:4046`): Loading a module chosen at runtime (from a variable) can pull in and run attacker-influenced code paths. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.
Evidence: try: __import__(pkg) deps[pkg] = True except ImportError: deps[pkg] =
Fix: Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.
Location: server cohort/server.py
Tools 12
Each tool and what it can reach — statically extracted from the published source.
cohort_errorno sensitive capabilitycohort_find_agentsno sensitive capabilitycohort_partnership_graphno sensitive capabilitycohort_postno sensitive capabilitycohort_post_and_continueno sensitive capabilitycohort_readyno sensitive capabilitycohort_respondno sensitive capabilitycohort_route_taskno sensitive capabilitydesktop_actionno sensitive capabilitydesktop_statusno sensitive capability
Show 2 more tools ↓
local_llm_generateno sensitive capabilitylocal_llm_modelsno sensitive capability
What this scan could not see
- Tools were statically extracted from the published source (12 recovered), not enumerated from a running server. Tool-poisoning, Unicode-smuggling, capability and toxic-flow analysis ran on this inferred surface, but a mis-parsed registration could be missed or mis-attributed, so tool-derived findings are capped below “confirmed”. To grade the real runtime surface, scan the running server: --command "npx -y <package>".
Versions 1
Scan history per published version. The engine is deterministic — the same version always yields the same score, so a changed score means the package itself changed.
| Version | Score | Findings | Engine | Scanned |
|---|---|---|---|---|
v0.4.47 latest |
A 93/100 | 12 | 1.8.0 | 2026-07-23 |
Embed this score
Show this server's live Trust Score in your README, docs or website. The badge is served straight from the registry and updates automatically after every rescan — no API key needed. It links back to this page, so anyone who sees the grade can also read the findings behind it instead of taking a number on faith.
Verify this score yourself
The score above is reproducible: the same package version always yields the same result. Run it locally or over the free API — no account, no LLM, fully deterministic.
npx mcptrustchecker scan cohort --online --registry pypiMore in Developer Tools
Security scan results for the 1stay MCP server.
Security scan results for the Adguard Home MCP server.
Security scan results for the Ado Browser MCP server.
Security scan results for the Adobe Experience Dev MCP server.
Security scan results for the Aemet MCP server.
Security scan results for the Affinity Mcp Bridge MCP server.