TITAN — TypeScript AI agent framework with multi-agent orchestration, provider routing, approval gates, receipts, evals, gateway APIs, and a React Mission Control dashboard. Open-source, MIT licensed.
- capability exposureinferred+35
- recent driftinferred+20
- tool safetyinferred+24
- supply-chainattested+6
- trust mitigatorsmixed−3
attestedinferredmixed
The A–E grade is our heuristic synthesis — a "review this" prompt, not a verdict. Each factor is tagged by what backs it: attested (a verifiable record), reported (a third party's claim), or inferred (our own heuristic, e.g. permissions). See methodology.
graded 14m ago · see ecosystem CVEs →
no known CVEs for this server.
- highdangerous code
dynamic exec: vm exec, new Function() · credential logged in 1 file(s)
- highhidden prompt content
1 file(s) with hidden prompt content: package/assets/agent-skills/browser-testing-with-devtools/SKILL.md (skill-exfil): "secret→sink: - **Read-only by default.** Use JavaScript ex…
analyzed v6.5.3 · analyzer v17 · 1d ago
skills & prompt files 22
- ⚠ hidden: skill: skill-exfilpackage/assets/agent-skills/browser-testing-with-devtools/SKILL.md:72
secret→sink: - **Read-only by default.** Use JavaScript execution for inspecting state (reading variables, querying the DOM, checking computed values), not for modifying page behavior.
- skillpackage/assets/agent-skills/api-and-interface-design/SKILL.md
- skillpackage/assets/agent-skills/ci-cd-and-automation/SKILL.md
- skillpackage/assets/agent-skills/code-review-and-quality/SKILL.md
- skillpackage/assets/agent-skills/code-simplification/SKILL.md
- skillpackage/assets/agent-skills/context-engineering/SKILL.md
- skillpackage/assets/agent-skills/debugging-and-error-recovery/SKILL.md
- skillpackage/assets/agent-skills/deprecation-and-migration/SKILL.md
- skillpackage/assets/agent-skills/documentation-and-adrs/SKILL.md
- skillpackage/assets/agent-skills/doubt-driven-development/SKILL.md
- skillpackage/assets/agent-skills/frontend-ui-engineering/SKILL.md
- skillpackage/assets/agent-skills/git-workflow-and-versioning/SKILL.md
- skillpackage/assets/agent-skills/idea-refine/SKILL.md
- skillpackage/assets/agent-skills/incremental-implementation/SKILL.md
- skillpackage/assets/agent-skills/performance-optimization/SKILL.md
- skillpackage/assets/agent-skills/planning-and-task-breakdown/SKILL.md
- skillpackage/assets/agent-skills/security-and-hardening/SKILL.md
- skillpackage/assets/agent-skills/shipping-and-launch/SKILL.md
- skillpackage/assets/agent-skills/source-driven-development/SKILL.md
- skillpackage/assets/agent-skills/spec-driven-development/SKILL.md
- skillpackage/assets/agent-skills/test-driven-development/SKILL.md
- skillpackage/assets/agent-skills/using-agent-skills/SKILL.md
danger signals7
- dynamic code executionvm execpackage/dist/skills/registry.js:632
const result = await vm.runInNewContext(wrapped, sandbox, { timeout: 1e4 }); - dynamic code executionnew Function()package/ui/dist/assets/TitanCanvas-TuYZSZO7.js:276
const fn = new Function('React', 'ReactDOM', 'titan', code + '\\n' + returnChecks); - suspicious endpointapi.telegram.orgpackage/dist/cli/onboard.js:111
const res = await fetch(`https://api.telegram.org/bot${token}/getMe`, { signal: ctl }); - over-broad OAuth scopehttps://www.googleapis.com/auth/drivepackage/dist/config/schema.js:433
"https://www.googleapis.com/auth/drive", - over-broad OAuth scopehttps://www.googleapis.com/auth/documentspackage/dist/config/schema.js:434
"https://www.googleapis.com/auth/documents", - over-broad OAuth scopehttps://www.googleapis.com/auth/spreadsheetspackage/dist/config/schema.js:435
"https://www.googleapis.com/auth/spreadsheets", - credential in logscredential in logpackage/dist/cli/index.js:464
console.log(` ${chalk.white(skill.name)} ${chalk.gray(`v${skill.version}`)} \u2014 ${apiKey}`);
- recent drift+20 capability drift →
- supply-chain+6 supply-chain hub →
[](https://mcpobservatory.com/servers/npm:titan-agent/security) Heuristic, inferred signals — false positives (legitimately powerful tools, forks, language ports) are expected. Treat each as "review this", not a verdict. See the ecosystem-wide picture on the security hub, or the fleet security of djtony707.