← ClaudeAtlas

tech-stack-analyzerlisted

Orchestrator-side entry point for "build me X"-style tasks. Translates a user's task description into a concrete spawn plan — stack + required skills + suggested personas + missing-skill bootstrap prompts. Always pauses for user redirect before spawning the team.
shashankcm95/claude-power-loom · ★ 2 · AI & Automation · score 68
Install: claude install-skill shashankcm95/claude-power-loom
# Tech-Stack Analyzer (HETS H.2.5) The orchestrator-side entry point for "build me X"-style tasks. Translates a user's task description into a concrete spawn plan: stack + required skills + suggested personas + missing-skill bootstrap prompts. Always pauses for user redirect before spawning the team. Implements the [tech-stack-analyzer](../agent-team/patterns/tech-stack-analyzer.md) and [skill-bootstrapping](../agent-team/patterns/skill-bootstrapping.md) patterns. ## When to use this skill Trigger when: - User asks for help building a new project ("I want to build a marketing site", "spin up a Java service for X") - User asks for help on a project where the stack is unspecified - Periodic catalog audits ("what skills would we need to handle the kinds of tasks I keep doing?") **Skip** when: - Task is pre-specified (e.g., chaos-test runs, scheduled release-pipeline runs — they have fixed personas) - User explicitly says "just spawn the standard team, don't analyze" - Task is a one-off question, not a multi-step build ## Workflow (5 steps + 2 user gates) ### Step 1 — Parse user intent Extract from the user's prose: - **Intent** — `build` | `refactor` | `debug` | `audit` | `migrate` | `evaluate` - **Domain** — `web` | `mobile` | `backend` | `data` | `ml` | `infra` | `security` - **Constraints** (if any) — deadline, scale, compliance, team-size, cost ceiling If task is too vague to extract intent + domain → emit ONE clarifying question (not a series). Don't spawn until the