io.github.no-human-ai/no_human

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

  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/no_human/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: downloads and pipes straight to a shell (src/no_human/cli/init_cmd.py: pipes a download into a shell). HIGH: executes content derived from a network response (src/no_human/vcs/git.py: passes a network response into an exec/eval call). has both shell/exec and destructive filesystem writes (src/no_human/config.py: subprocess.* call; src/no_human/config.py: destructive filesystem call (shutil.rmtree/os.remove/etc.)). discloses shell/exec capability: 46 files (src/no_human/config.py, src/no_human/docs_gen.py, src/no_human/doctor.py, +43 more) discloses broad/destructive filesystem access: 18 files (src/no_human/config.py, src/no_human/updates.py, src/no_human/api/app.py, +15 more) Capability DISCLOSURE, not a verdict: static analysis sees the primitive is present and reachable, not whether its use is attacker-controlled.

• Version behavior

  fail10% of grade

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

  vs prior 0.1.4 (minor bump): new capability primitive(s) in source: child_process, dynamic-eval, fs-write, http-client, net, subprocess; new hardcoded outbound host/BCC destination(s): 127.0.0.1, acme.atlassian.net, adaptivecards.io, api.github.com, api.linear.app, api.monday.com, api.slack.com, app.circleci.com, astral.sh, attacker, attacker.example, build.example.com, circleci.com, code.example.com, console.anthropic.com, developer.monday.com, developers.openai.com, docs.github.com, docs.gitlab.com, evil.com, getnohuman.com, git-scm.com, github.com, graph.microsoft.com, host, id.atlassian.com, learn.microsoft.com, linear.app, localhost.evil.com, outlook.office.com, platform.claude.com, pypi.org, python.org, support.atlassian.com, us-assets.i.posthog.com, us.i.posthog.com, user, www.jenkins.io, xxxxx.logic.azure.com.

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