io.github.ariffazil/arifos-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

  pass25% of grade

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

  repository linkage verified

• Credential hygiene

  warn15% of grade

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

  core/kernel/init_000_anchor.py: assigns a literal to secret (arif…et (len 28)) — non-provider secret literal; prefer environment intake. Reads secrets from the environment (e.g. aaa_mcp/__main__.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: aclip_cai/tools/safety_guard.py: references a sensitive credential path (id_rsa) — the tool-poisoning exfiltration target. HIGH: shell/exec capability AND network egress together — download-and-execute risk (aaa_mcp/integrations/container_controller.py: subprocess.* call; aaa_mcp/capabilities/t6_web_search.py: outbound HTTP/socket call). HIGH: shell/exec capability AND destructive filesystem writes together (aaa_mcp/integrations/container_controller.py: subprocess.* call; aaa_mcp/streamable_http_server.py: open(…, 'w'/'a'/'x') write-mode file access). discloses shell/exec capability: aaa_mcp/integrations/container_controller.py: subprocess.* call discloses shell/exec capability: aaa_mcp/integrations/mcp_container_tools.py: subprocess.* call discloses shell/exec capability: aaa_mcp/wrappers/opa_policy.py: subprocess.* call discloses shell/exec capability: aclip_cai/console_tools.py: subprocess.* call discloses shell/exec capability: aclip_cai/tools/system_monitor.py: subprocess.* call discloses shell/exec capability: core/physics/thermodynamics.py: subprocess.* call discloses shell/exec capability: scripts/deploy_production.py: subprocess.* call discloses shell/exec capability: tests/test_skill_capture_terminal_output.py: subprocess.* call discloses shell/exec capability: tests/test_skill_route_tasks.py: subprocess.* call discloses broad/destructive filesystem access: aaa_mcp/streamable_http_server.py: open(…, 'w'/'a'/'x') write-mode file access discloses broad/destructive filesystem access: aaa_mcp/wrappers/opa_policy.py: destructive filesystem call (shutil.rmtree/os.remove/etc.) discloses broad/destructive filesystem access: core/telemetry.py: open(…, 'w'/'a'/'x') write-mode file access discloses broad/destructive filesystem access: scripts/deploy_production.py: open(…, 'w'/'a'/'x') write-mode file access discloses broad/destructive filesystem access: scripts/housekeeping.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.ariffazil/[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.

  manifest/source references tracking behavior — disclosure