io.github.theoddden/terradev

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_drift_auth_and_errors.py: assigns a literal to token (azur…en (len 15)) — non-provider secret literal; prefer environment intake. Reads secrets from the environment (e.g. terradev_cli/cli.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: terradev_cli/providers/crusoe_provider.py reads a credential path, executes commands, and makes network calls in the same file — the combination that makes exfiltration possible. references credential paths: terradev_cli/providers/crusoe_provider.py: references a credential path (/.ssh/) installer script pipes a download into a shell — conventional, but worth reading before running (terradev_cli/terraform/modules/gpu-node-aws/bootstrap.sh: pipes a download into a shell). has both shell/exec and network egress — capable of download-and-run, though no such path was observed (terradev_cli/commands/agent_infra/dependency_manager.py: subprocess.* call; terradev_cli/cli_hf_spaces.py: outbound HTTP/socket call). has both shell/exec and destructive filesystem writes (terradev_cli/commands/agent_infra/dependency_manager.py: subprocess.* call; terradev_cli/cli.py: open(…, 'w'/'a'/'x') write-mode file access). discloses shell/exec capability: 42 files (terradev_cli/commands/agent_infra/dependency_manager.py, terradev_cli/commands/agent_infra/landlock.py, terradev_cli/commands/inference.py, +39 more) discloses dynamic code execution: terradev_cli/commands/mlops.py: uses eval()/exec() discloses broad/destructive filesystem access: 53 files (terradev_cli/cli.py, terradev_cli/commands/_api.py, terradev_cli/commands/agent_infra/sandbox.py, +50 more) Capability DISCLOSURE, not a verdict: static analysis sees the primitive is present and reachable, not whether its use is attacker-controlled.

• Version behavior

  pass10% of grade

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

  tool surface, capabilities, install scripts, and publisher unchanged vs prior version 6.2.8 (patch bump); no rug-pull drift.

• 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 (Datadog) — disclosure, verify opt-in; manifest/source references analytics behavior — disclosure