io.github.zrk222/code-factory

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_evidence_supply_line.py: assigns a literal to secret (do-n…nd (len 26)) — non-provider secret literal; prefer environment intake. Reads secrets from the environment (e.g. factoryline/assurance.py) — the recommended intake shape.

• Permission scope

  warn15% of grade

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

  references credential paths: factoryline/mcp_setup.py: references a credential path (/mcp.json) has both shell/exec and network egress — capable of download-and-run, though no such path was observed (setup.py: subprocess.* call; factoryline/hosted_identity.py: outbound HTTP/socket call). has both shell/exec and destructive filesystem writes (setup.py: subprocess.* call; factoryline/agent_proof_bridge.py: destructive filesystem call (shutil.rmtree/os.remove/etc.)). discloses shell/exec capability: 13 files (setup.py, factoryline/assembly.py, factoryline/assurance.py, +10 more) discloses broad/destructive filesystem access: 25 files (factoryline/agent_proof_bridge.py, factoryline/appforge_eas.py, factoryline/appforge_oracle.py, +22 more) 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 0.46.0 (patch bump): new capability primitive(s) in source: fs-write, http-client, subprocess; new hardcoded outbound host/BCC destination(s): 127.0.0.1, api.github.com, api.provider-a.example, api.provider-b.example, bootstrap.test, code-factory.dev, developer.apple.com, docs.oasis-open.org, example.com, example.invalid, example.test, github.com, id.acme.test, id.example, id.example.test, id.integration.test, id.other.test, in-toto.io, issuer.example, issuer.example.test, opencode.ai, plugins.jetbrains.com, provider-a.example, pypi.org, raw.githubusercontent.com, slsa.dev, static.modelcontextprotocol.io, token.actions.githubusercontent.com, www.w3.org.

• Transport config

  fail10% of grade

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

  source performs runtime self-update / download-and-execute (bypasses version_behavior); manifest/source references analytics behavior — disclosure