The Lead Stack: Agent-Agnostic WorkflowsBLOCK
AI agent workflows for cross-functional teams. Tech Leads, PMs, and HR can run RTK skills via IDE or Web App.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
intent-brief → spec → plan → diff → review-report → release
[](https://github.com/bronz3beard/ai.tech-lead-stack/actions/workflows/ci.yml) [](https://github.com/bronz3beard/ai.tech-lead-stack/actions/workflows/codeql.yml) [](https://scorecard.dev/viewer/?uri=github.com/bronz3beard/ai.tech-lead-stack) [](https://www.bestpractices.dev/projects/14956) [](https://www.bestpractices.dev/projects/14956) [](LICENSE) [](package.json) [](https://www.npmjs.com/package/tech-lead-stack) [](https://www.npmjs.com/package/tech-lead-stack) [](https://bundlephobia.com/package/tech-lead-stack) [](docs/skills.md) [](docs/running-the-mcp-server.md) [](docs/agent-setup.md) [![Ed
70c769af1252OBSERVED · 2026-10-11Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add tech-lead-stack --env ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY} --env AUTH_SECRET=${AUTH_SECRET} --env ENCRYPTION_KEY=${ENCRYPTION_KEY} --env GEMINI_API_KEY=${GEMINI_API_KEY} -- npx -y @zenithfoundry/[email protected]{
"mcpServers": {
"tech-lead-stack": {
"command": "npx",
"args": [
"-y",
"@zenithfoundry/[email protected]"
],
"env": {
"ANTHROPIC_API_KEY": "${ANTHROPIC_API_KEY}",
"AUTH_SECRET": "${AUTH_SECRET}",
"ENCRYPTION_KEY": "${ENCRYPTION_KEY}",
"GEMINI_API_KEY": "${GEMINI_API_KEY}"
}
}
}
}Exposed tools (18)
14 read · 4 write · 0 destructive.
| Tool | Risk | Description |
|---|---|---|
apply_patch | write | Applies minimal SEARCH/REPLACE blocks to a file instead of rewriting it completely. |
approve_knowledge_item | read | Sets the approval state of a Knowledge Item. |
code_search | read | Semantic search across the repository code using local embeddings. |
create_knowledge_item | write | Creates or updates a Knowledge Item from the current context. |
get_skill | read | CORE EXECUTION: Reads a specific skill (e.g. |
get_skills | read | Reads the content of one or more skill markdown files. |
list_knowledge_items | read | Lists all available Antigravity Knowledge Items. |
list_skills | write | MANDATORY START: Lists ALL available Tech-Lead Stack architecture skills and workflows. |
plan_pipeline | read | Returns the ordered phase->skill chain for a task. |
read_knowledge_item | read | Reads a specific Knowledge Item and its artifacts. |
read_region | read | Reads a specific line range from a file. |
reflexion_loop | read | ✨ SPECIAL FEATURE: Self-correcting plan loop. Gemini drafts an implementation plan, Claude grades it 0-10 against the Four Pillars and returns one fix, repeat until it passes or caps. Requires GEMINI_API_KEY + ANTHROPIC_API_KEY. |
reflexion_loop_sub_max | read | [DEV-TEAM · SUB-MAX] Plan hardening for $100/mo tier. |
reflexion_loop_sub_pro | read | [DEV-TEAM · SUB-PRO] Plan hardening for $20/mo tier. |
reflexion_resume | read | Resumes a parked Reflexion Loop. |
reflexion_status | write | Checks the status of an asynchronous Reflexion Loop run. |
repo_map | read | Retrieves a compact map of the repository files and top-level symbol signatures. |
verify_mission_alignment | read | MANDATORY PRE-FLIGHT: Validates that the agent is aligned with the operational boundaries and has initialized the required mission skills. |
Trust audit
BLOCKgrade F · trust 53/100 Do not install this without reading the findings. The audit found something that could harm you or your machine.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | FAIL |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (8 observation(s))
- Network
- declared (12 observation(s))
- Shell
- declared (6 observation(s))
- Dependencies
- not all pinned
- Secrets in source
- found
Findings (25)
parsedYaml = yaml.load(yamlBlock);
'Scans skills, scripts, and inputs for malicious patterns (`curl \\| bash`, `eval()`).',
return { rejectUnauthorized: false };DATABASE_URL="postgresql://postgres:password@localhost:5432/agent_dashboard?schema=public"
'postgresql://postgres:postgres@localhost:5432/postgres',
'Cross-platform security scanner detecting malware, prompt injection, and exfiltration.',
`curl -s -o /dev/null -w '%{http_code}' http://127.0.0.1:${port}/`,`curl -i -L -s http://localhost:${port} || curl -i -L -s http://127.0.0.1:${port}`,env: { DATABASE_URL: 'postgresql://u:p@h/db' },DATABASE_URL: 'postgresql://u:p@h/db',
'postgresql://u:[email protected]/db',
.markdownlint.json
.prettierignore
.release-please-manifest.json
import rootConfig from '../../eslint.config.mjs';
'^@zenithfoundry/tech-lead-stack/db$': '<rootDir>/../../packages/core/src/lib/prisma.ts',
'^@zenithfoundry/tech-lead-stack/crypto$': '<rootDir>/../../packages/core/src/lib/crypto.ts',
'^@zenithfoundry/tech-lead-stack/telemetry-service$': '<rootDir>/../../packages/core/src/lib/telemetry-service.ts',
'^@zenithfoundry/tech-lead-stack/trace-utils$': '<rootDir>/../../packages/core/src/lib/trace-utils.ts',
rejectUnauthorized: false,
@ai-sdk/react, @base-ui/react, @next-auth/prisma-adapter, @prisma/adapter-pg, @prisma/client, @radix-ui/react-slot, @tailwindcss/postcss, @tailwindcss/typography
@types/bcryptjs, @types/pg, @types/react-syntax-highlighter, dotenv, p-map, postcss, tailwind-merge, tailwindcss
@ai-sdk/anthropic, @ai-sdk/google, @ai-sdk/openai, @modelcontextprotocol/sdk, @prisma/adapter-pg, @prisma/client, adm-zip, dotenv
instructions," "bypass safety," or "disregard guidelines."
Ensure the `projectId` sent by the GitHub Action matches the `id` of the project in your database. Check the server logs for `[POST] /api/design-review/session` errors.
Gates applied: no_behavioural_pass.
70c769af1252full audit observations/trust-audit/mcp-server/bronz3beard__ai.tech-lead-stack.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-11 | 70c769af1252 | BLOCK | F | 53 | first audit |
Questions
What is the ai.tech-lead-stack MCP server?
AI agent workflows for cross-functional teams. Tech Leads, PMs, and HR can run RTK skills via IDE or Web App.
What tools does ai.tech-lead-stack expose?
18 in total: 14 read-only, 4 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.
Is ai.tech-lead-stack safe to connect to an agent?
No — not without reading the findings first. The audit graded it F (53/100) and found 5 critical or high issues in the source. Each one is listed on this page with the file and line it is on.
What credentials does ai.tech-lead-stack need?
It reads ANTHROPIC_API_KEY, AUTH_SECRET, ENCRYPTION_KEY, GEMINI_API_KEY, GEMINI_API_KEY_PRECEDENCE, GITHUB_SECRET, GITHUB_TOKEN, GOOGLE_GENERATIVE_AI_API_KEY, JULES_API_KEY, NEXTAUTH_SECRET, NEXTAUTH_URL and OPENAI_API_KEY from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How does ai.tech-lead-stack run?
It speaks stdio, so it runs as a local process your client starts. It is published on npm as @zenithfoundry/tech-lead-stack at 2.2.0.
How current is this page?
The grade is for one exact copy of the source (70c769af1252), read on 2026-10-11. The repository is watched and re-audited when it changes.