io.github.Creed-Space/mettle-mcp

What we checked

Six static checks, weighted by risk. Every result reflects only what could be observed in the published package and repository — never intent.

• Injection surface

  unscannable25% of grade

  Tool descriptions/manifest scanned for instruction-injection patterns (imperatives at the model, hidden text, 'ignore previous', data-exfil URLs).

  No tool descriptions, server instructions, prompts, or resources were fetched; nothing to scan for injection.

• Supply chain

  fail25% of grade

  Package provenance: namespace verification, repo linkage, maintainer count, account age, postinstall scripts, typosquat distance.

  Python build script executes at build time: network call in build script

• Credential hygiene

  warn15% of grade

  How the server takes secrets (env vs plaintext config vs hardcoded); secrets appearing in tool schemas.

  tests/test_challenge_adapter.py: assigns a literal to token (VCP:…:4 (len 37)) — non-provider secret literal; prefer environment intake. tests/test_challenge_adapter.py: assigns a literal to token (VCP:…:2 (len 44)) — non-provider secret literal; prefer environment intake. tests/test_challenge_adapter.py: assigns a literal to token (VCP:…lt (len 45)) — non-provider secret literal; prefer environment intake. tests/test_challenge_adapter.py: assigns a literal to token (VCP:…:5 (len 51)) — non-provider secret literal; prefer environment intake. tests/test_challenge_adapter.py: assigns a literal to token (VCP:…:5 (len 40)) — non-provider secret literal; prefer environment intake. tests/test_challenge_adapter.py: assigns a literal to token (VCP:…:2 (len 38)) — non-provider secret literal; prefer environment intake. tests/test_database.py: assigns a literal to api_key (lega…ey (len 17)) — non-provider secret literal; prefer environment intake. tests/test_main_coverage.py: assigns a literal to token (sign…en (len 12)) — non-provider secret literal; prefer environment intake. tests/test_main_coverage.py: assigns a literal to token (reco…en (len 15)) — non-provider secret literal; prefer environment intake. tests/test_main_coverage.py: assigns a literal to secret (wron…es (len 42)) — non-provider secret literal; prefer environment intake. tests/test_p3_coverage_gaps.py: assigns a literal to token (VCP:…al (len 48)) — non-provider secret literal; prefer environment intake. tests/test_p3_coverage_gaps.py: assigns a literal to token (VCP:…er (len 34)) — non-provider secret literal; prefer environment intake. tests/test_p3_coverage_gaps.py: assigns a literal to token (VCP:…er (len 22)) — non-provider secret literal; prefer environment intake. tests/test_security_features.py: assigns a literal to api_key (batc…ey (len 13)) — non-provider secret literal; prefer environment intake. tests/test_vcp.py: assigns a literal to token (VCP:…on (len 38)) — non-provider secret literal; prefer environment intake. tests/test_vcp.py: assigns a literal to token (VCP:…or (len 26)) — non-provider secret literal; prefer environment intake. tests/test_vcp.py: assigns a literal to token (VCP:…gh (len 31)) — non-provider secret literal; prefer environment intake. tests/test_vcp.py: assigns a literal to token (VCP:…0Z (len 58)) — non-provider secret literal; prefer environment intake. tests/test_vcp.py: assigns a literal to token (VCP:…st (len 42)) — non-provider secret literal; prefer environment intake. tests/test_vcp.py: assigns a literal to token (VCP:….0 (len 27)) — non-provider secret literal; prefer environment intake. tests/test_vcp.py: assigns a literal to token (VCP:…lp (len 49)) — non-provider secret literal; prefer environment intake. tests/test_vcp.py: assigns a literal to token (VCP:….0 (len 31)) — non-provider secret literal; prefer environment intake. tests/test_vcp_coverage.py: assigns a literal to token (VCP:…al (len 54)) — non-provider secret literal; prefer environment intake. tests/test_vcp_coverage.py: assigns a literal to token (VCP:…gh (len 34)) — non-provider secret literal; prefer environment intake. tests/test_vcp_integration.py: assigns a literal to token (VCP:…al (len 63)) — non-provider secret literal; prefer environment intake. Reads secrets from the environment (e.g. mettle/_http.py) — the recommended intake shape.

• Permission scope

  warn15% of grade

  Declared tools vs. breadth (filesystem, network, exec); flags shell-exec and unbounded filesystem access.

  has both shell/exec and network egress — capable of download-and-run, though no such path was observed (mettle/holder.py: subprocess.* call; mettle/_http.py: outbound HTTP/socket call). discloses shell/exec capability: mettle/holder.py: subprocess.* call discloses dynamic code execution: mettle/legacy_session_store.py: uses eval()/exec() Capability DISCLOSURE, not a verdict: static analysis sees the primitive is present and reachable, not whether its use is attacker-controlled.

• Version behavior

  fail10% of grade

  Diff of tool definitions between versions; new permissions or changed descriptions in a patch release (the postmark-mcp class).

  vs prior 0.4.8 (minor bump): new capability primitive(s) in source: dynamic-eval, fs-write, http-client, net, subprocess; new hardcoded outbound host/BCC destination(s): 10.0.0.1, 127.0.0.1, 169.254.169.254, 93.184.216.34, a.com, api.render.com, app.creed.space, attacker.example, b.com, c.com, del.com, evil.example, example.com, example.test, external-host.com, files.pythonhosted.org.example, github.com, github.example, hook.com, json-schema.org, mettle.example, mettle.sh, new.com, old.com, only.com, other.io, pypi.org, service.internal, testserver, vault.example, www.cloudflare.com, x.com.

• Transport config

  warn10% of grade

  Remote servers: TLS and auth mode (none/token/OAuth). Local servers: whether the manifest indicates it phones home.

  manifest/source references tracking behavior — disclosure