io.github.sassyconsultingllc/sassymcp

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

  fail15% of grade

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

  tools/verify_audit_fixes.py: embeds a PRIVATE KEY block. sassymcp/modules/setup_wizard.py: assigns a literal to token (ghp_…re (len 19)) — non-provider secret literal; prefer environment intake. Reads secrets from the environment (e.g. hermes_node.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: sassymcp/modules/_security.py: references a sensitive credential path (/.ssh/) — the tool-poisoning exfiltration target. HIGH: sassymcp/modules/linux.py: references a sensitive credential path (/.ssh/) — the tool-poisoning exfiltration target. HIGH: tools/verify_audit_fixes.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 (sassymcp/_phone_status.py: subprocess.* call; hermes_node.py: outbound HTTP/socket call). HIGH: shell/exec capability AND destructive filesystem writes together (sassymcp/_phone_status.py: subprocess.* call; sassymcp/_atomic.py: destructive filesystem call (shutil.rmtree/os.remove/etc.)). discloses shell/exec capability: sassymcp/_phone_status.py: subprocess.* call discloses shell/exec capability: sassymcp/_platform.py: subprocess.* call discloses shell/exec capability: sassymcp/auth.py: subprocess.* call discloses shell/exec capability: sassymcp/server.py: subprocess.* call discloses shell/exec capability: sassymcp/supervisor.py: subprocess.* call discloses shell/exec capability: sassymcp/desktop/bridge.py: subprocess.* call discloses shell/exec capability: sassymcp/modules/app_launcher.py: subprocess.* call discloses shell/exec capability: sassymcp/modules/linux.py: subprocess.* call discloses shell/exec capability: sassymcp/modules/selfmod.py: subprocess.* call discloses shell/exec capability: sassymcp/modules/web_inspector.py: subprocess.* call discloses shell/exec capability: sassymcp/overlay/mesh.py: subprocess.* call with shell=True (command-injection prone) discloses shell/exec capability: sassymcp-vscode/out/cockpitData.js: imports child_process discloses shell/exec capability: sassymcp-vscode/out/hermes.js: imports child_process discloses shell/exec capability: sassymcp-vscode/out/installer.js: imports child_process discloses shell/exec capability: sassymcp-vscode/out/setupWizard.js: imports child_process discloses shell/exec capability: sassymcp-vscode/src/cockpitData.ts: imports child_process and calls exec/spawn discloses shell/exec capability: sassymcp-vscode/src/hermes.ts: imports child_process and calls exec/spawn discloses shell/exec capability: sassymcp-vscode/src/installer.ts: imports child_process and calls exec/spawn discloses shell/exec capability: sassymcp-vscode/src/setupWizard.ts: imports child_process and calls exec/spawn discloses shell/exec capability: tests/test_atomic_write.py: subprocess.* call discloses shell/exec capability: tests/test_audit_io.py: subprocess.* call discloses shell/exec capability: tests/test_concurrency_integration.py: subprocess.* call discloses shell/exec capability: tests/test_concurrency_stress.py: subprocess.* call discloses shell/exec capability: tests/test_supervisor_lifecycle.py: subprocess.* call discloses dynamic code execution: sassymcp/_jobctl.py: uses eval/exec/__import__ discloses dynamic code execution: sassymcp/server.py: uses eval/exec/__import__ discloses dynamic code execution: sassymcp/modules/persona.py: uses eval/exec/__import__ discloses dynamic code execution: sassymcp/modules/setup_wizard.py: uses eval/exec/__import__ discloses dynamic code execution: sassymcp/modules/tools_manager.py: uses eval/exec/__import__ discloses dynamic code execution: tools/verify_audit_fixes.py: uses eval/exec/__import__ discloses broad/destructive filesystem access: sassymcp/_atomic.py: destructive filesystem call (shutil.rmtree/os.remove/etc.) discloses broad/destructive filesystem access: sassymcp/_audit_io.py: open(…, 'w'/'a'/'x') write-mode file access discloses broad/destructive filesystem access: sassymcp/server.py: destructive filesystem call (shutil.rmtree/os.remove/etc.) discloses broad/destructive filesystem access: sassymcp/modules/_security.py: path traversal ('../') combined with file access discloses broad/destructive filesystem access: sassymcp/modules/fileops.py: open(…, 'w'/'a'/'x') write-mode file access discloses broad/destructive filesystem access: sassymcp/modules/fileops.py: destructive filesystem call (shutil.rmtree/os.remove/etc.) discloses broad/destructive filesystem access: sassymcp/modules/github_quick.py: path traversal ('../') combined with file access discloses broad/destructive filesystem access: sassymcp/modules/setup_wizard.py: destructive filesystem call (shutil.rmtree/os.remove/etc.) discloses broad/destructive filesystem access: sassymcp/modules/setup_wizard.py: path traversal ('../') combined with file access discloses broad/destructive filesystem access: sassymcp/modules/shell.py: destructive filesystem call (shutil.rmtree/os.remove/etc.) discloses broad/destructive filesystem access: sassymcp/modules/updater.py: open(…, 'w'/'a'/'x') write-mode file access discloses broad/destructive filesystem access: tests/test_intercept.py: path traversal ('../') combined with file access discloses broad/destructive filesystem access: tools/verify_audit_fixes.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.sassyconsultingllc/[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); sends telemetry to third-party SDK (google-analytics) — disclosure, verify opt-in; manifest/source references analytics behavior — disclosure