io.github.aidongise-cell/prism-scanner

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

  pass25% of grade

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

  Python build script has no build-time network/exec; repository linkage verified

• Credential hygiene

  warn15% of grade

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

  tests/test_scanner.py: assigns a literal to token (a3f8…gI (len 34)) — non-provider secret literal; prefer environment intake. Reads secrets from the environment (e.g. src/prism/engines/ast_engine.py) — the recommended intake shape.

• Permission scope

  fail15% of grade

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

  HIGH: src/prism/engines/ast_engine.py: references a sensitive credential path (/.ssh/) — the tool-poisoning exfiltration target. HIGH: tests/test_scanner.py: references a sensitive credential path (/.ssh/) — the tool-poisoning exfiltration target. HIGH: shell/exec capability AND destructive filesystem writes together (src/prism/cleaner.py: subprocess.* call; src/prism/cleaner.py: destructive filesystem call (shutil.rmtree/os.remove/etc.)). discloses shell/exec capability: src/prism/cleaner.py: subprocess.* call discloses shell/exec capability: src/prism/engines/ast_engine.py: subprocess.* call with shell=True (command-injection prone) discloses shell/exec capability: src/prism/engines/residue_engine.py: subprocess.* call discloses shell/exec capability: src/prism/fetcher.py: subprocess.* call discloses dynamic code execution: src/prism/engines/pattern_engine.py: uses eval/exec/__import__ discloses dynamic code execution: tests/test_scanner.py: uses eval/exec/__import__ discloses broad/destructive filesystem access: src/prism/cleaner.py: destructive filesystem call (shutil.rmtree/os.remove/etc.) discloses broad/destructive filesystem access: src/prism/engines/residue_engine.py: recursive glob ('**') from a broad root discloses broad/destructive filesystem access: src/prism/fetcher.py: destructive filesystem call (shutil.rmtree/os.remove/etc.) discloses broad/destructive filesystem access: tests/test_scanner.py: open(…, 'w'/'a'/'x') write-mode file access discloses broad/destructive filesystem access: tests/test_scanner.py: destructive filesystem call (shutil.rmtree/os.remove/etc.) Capability DISCLOSURE, not a verdict: static analysis sees the primitive is present and reachable, not whether its use is attacker-controlled.

• Version behavior

  unscannable10% of grade

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

  no prior version to diff (first sight of io.github.aidongise-cell/[email protected]); version-drift / rug-pull cannot be evaluated. Cold-start risk is covered by point-in-time checks (injection_surface, supply_chain, credential_hygiene), not here.

• 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 tracking behavior — disclosure