io.github.soufianetahiri/mception

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: os.system/popen in build script; no resolvable repository linked to the package (provenance unverifiable)

• Credential hygiene

  pass15% of grade

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

  Reads secrets from the environment (e.g. src/mception/config.py) — the recommended intake shape. PASS = no static red flag; static analysis cannot prove the code honors env-based secret handling at runtime.

• Permission scope

  fail15% of grade

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

  HIGH: examples/demo_evil/evil.py: references a sensitive credential path (/.ssh/) — the tool-poisoning exfiltration target. HIGH: examples/demo_full/server.py: references a sensitive credential path (/.ssh/) — the tool-poisoning exfiltration target. HIGH: src/mception/rules/text_rules.py: references a sensitive credential path (\.ssh/) — the tool-poisoning exfiltration target. HIGH: tests/test_audit_end_to_end.py: references a sensitive credential path (/.ssh/) — the tool-poisoning exfiltration target. HIGH: tests/test_text_rules.py: references a sensitive credential path (/.ssh/) — the tool-poisoning exfiltration target. HIGH: shell/exec capability AND network egress together — download-and-execute risk (examples/demo_full/server.py: subprocess.* call with shell=True (command-injection prone); examples/demo_full/server.py: outbound HTTP/socket call). HIGH: shell/exec capability AND destructive filesystem writes together (examples/demo_full/server.py: subprocess.* call with shell=True (command-injection prone); packaging/build_bundle.py: destructive filesystem call (shutil.rmtree/os.remove/etc.)). discloses shell/exec capability: examples/demo_full/server.py: subprocess.* call with shell=True (command-injection prone) discloses shell/exec capability: examples/demo_full/setup.py: os.system/os.popen/os.exec* call discloses shell/exec capability: packaging/build_bundle.py: subprocess.* call discloses shell/exec capability: src/mception/engines/fetcher.py: subprocess.* call discloses shell/exec capability: src/mception/engines/sast.py: subprocess.* call discloses shell/exec capability: tests/test_baseline.py: subprocess.* call with shell=True (command-injection prone) discloses shell/exec capability: tests/test_code_rules.py: subprocess.* call with shell=True (command-injection prone) discloses shell/exec capability: tests/test_code_rules.py: os.system/os.popen/os.exec* call discloses shell/exec capability: tests/test_multilang.py: subprocess.* call with shell=True (command-injection prone) discloses shell/exec capability: tests/fixtures/servers/node_real_exec/src/index.js: imports child_process and calls exec/spawn discloses shell/exec capability: tests/fixtures/servers/node_regex_exec/src/index.js: imports child_process and calls exec/spawn discloses shell/exec capability: tests/fixtures/servers/python_subprocess/mcp_server.py: subprocess.* call with shell=True (command-injection prone) discloses dynamic code execution: src/mception/rules/node_rules.py: uses eval/exec/__import__ discloses dynamic code execution: src/mception/rules/supply_chain.py: uses eval/exec/__import__ discloses dynamic code execution: tests/test_code_rules.py: uses eval/exec/__import__ discloses dynamic code execution: tests/test_multilang.py: uses eval/exec/__import__ discloses dynamic code execution: tests/test_supply_chain.py: uses eval/exec/__import__ discloses dynamic code execution: tests/fixtures/servers/node_figma_plugin/code.js: uses eval(…) discloses dynamic code execution: tests/fixtures/servers/python_method_exec/mcp_server.py: uses eval/exec/__import__ discloses broad/destructive filesystem access: packaging/build_bundle.py: destructive filesystem call (shutil.rmtree/os.remove/etc.) discloses broad/destructive filesystem access: src/mception/server.py: destructive filesystem call (shutil.rmtree/os.remove/etc.) discloses broad/destructive filesystem access: src/mception/engines/dispatch.py: destructive filesystem call (shutil.rmtree/os.remove/etc.) discloses broad/destructive filesystem access: src/mception/engines/fetcher.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.soufianetahiri/[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

  warn10% of grade

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

  sends telemetry to third-party SDK (datadoghq) — disclosure, verify opt-in; manifest/source references tracking behavior — disclosure