Claude Code subagent imported from TheAstrelo/Claude-Pipeline (.claude/agents/README.md). Copyright stays with the author.

Interactive agents (helpers, not the engine)

Name: README
Author: TheAstrelo

These agents back the per-phase slash commands (/design, /ar, /security-review, …) when you run them interactively in Claude Code. The pipeline engine (run-pipeline.sh) does not use them — it inlines every phase prompt in build_prompt(). Editing an agent changes the interactive helper only, never engine behavior.

Reference-project content. Several of these prompts were written for the specific app the pipeline was extracted from (a Next.js + PostgreSQL SaaS called "RDO") and still name its tables, imports, and UI conventions as examples. Treat those as illustrations of shape, not rules for your project. Your real conventions belong in .claude/rules/*.md (see examples/reference-project-rules/ for a worked example). If you copy .claude/ into another project and use these agents interactively, adjust the project-specific references first.

{ "ocm": "1", "id": "theastrelo-claude-pipeline-readme-subagent", "kind": "agent", "name": "README", "description": "These agents back the per-phase slash commands (`/design`, `/ar`, `/security-review`, …) when you run them interactively in Claude Code. **The pipeline engine (`run-pipeline.sh`) does not use them** — it inlines every phase prompt in `build_prompt()`. Editing an agent changes the interactive helper only, never engine behavior.", "publisher": "TheAstrelo", "version": "1.0.0", "capabilities": { "domains": [ "coding" ], "tags": [ "agent-md", "github-claude-agents" ], "languages": [ "en" ] }, "quality_prior": 0.6, "examples": [ "These agents back the per-phase slash commands (`/design`, `/ar`, `/security-review`, …) when you run them interactively in Claude Code. **The pipeline engine (`run-pipeline.sh`) does not use them** — it inlines every phase prompt in `build_prompt()`. Editing an agent changes the interactive helper only, never engine behavior." ], "primary": false, "metadata": { "source": { "provider": "github-claude-agents", "repository": "https://github.com/TheAstrelo/Claude-Pipeline", "path": ".claude/agents/README.md", "ref": "909e9100cd3eda28c1485c45a98266d048e6dfce", "url": "https://github.com/TheAstrelo/Claude-Pipeline/blob/909e9100cd3eda28c1485c45a98266d048e6dfce/.claude/agents/README.md", "key": "TheAstrelo/Claude-Pipeline/.claude/agents/README.md" } }, "instructions": "# Interactive agents (helpers, not the engine)\n\nThese agents back the per-phase slash commands (`/design`, `/ar`,\n`/security-review`, …) when you run them interactively in Claude Code. **The\npipeline engine (`run-pipeline.sh`) does not use them** — it inlines every\nphase prompt in `build_prompt()`. Editing an agent changes the interactive\nhelper only, never engine behavior.\n\n**Reference-project content.** Several of these prompts were written for the\nspecific app the pipeline was extracted from (a Next.js + PostgreSQL SaaS\ncalled \"RDO\") and still name its tables, imports, and UI conventions as", "cost": { "context_tokens": 230 } }

README

About

Interactive agents (helpers, not the engine)

Use it

Manifest

Reviews