io.github.capad-xyz/searchts

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

  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. searchts/cli.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 (searchts/cli.py: pipes a download into a shell). has both shell/exec and destructive filesystem writes (searchts/cli.py: subprocess.* call; searchts/cli.py: destructive filesystem call (shutil.rmtree/os.remove/etc.)). discloses shell/exec capability: 4 files (searchts/cli.py, searchts/probe.py, searchts/transcribe.py, +1 more) discloses broad/destructive filesystem access: searchts/cli.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

  fail10% of grade

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

  vs prior 0.7.2 (minor bump): new capability primitive(s) in source: fs-write, http-client, subprocess; new hardcoded outbound host/BCC destination(s): api.exa.ai, api.github.com, api.groq.com, api.openai.com, api.search.brave.com, api.tavily.com, bad.test, capad.fyi, cdn.test, chromewebstore.google.com, cli.github.com, console.groq.com, d.test, deb.nodesource.com, demo.test, developer.mozilla.org, developers.cloudflare.com, en.wikipedia.org, exa.ai, example.com, fnm.vercel.app, fonts.googleapis.com, github.com, hnrss.org, instagram.com, mcp.exa.ai, news.ycombinator.com, nodejs.org, ok.test, platform.openai.com, r.jina.ai, raw.githubusercontent.com, redd.it, reddit.com, s, s.test, site.test, stackoverflow.com, tiktok.com, u, user, v.redd.it, vm.tiktok.com, www.bbc.co.uk, www.example.com, www.instagram.com, www.linkedin.com, www.reddit.com, www.tiktok.com, www.webshare.io, www.youtube.com, x.com, x.queue-it.net, x.test, youtu.be, youtube.com.

• 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 check for updates behavior — disclosure