diagram-architect¶
Turn a project's docs and code structure into an architecture diagram — it reads the repo model, builds a structured diagram of the real components and their relationships, and emits it in the requested format (draw.io .drawio XML by default, exported to SVG; Excalidraw fallback, mermaid, inline SVG, or Figma-MCP on request). Every element traces to the actual repo; nothing is invented.
At a glance¶
| Model | inherits the session's |
| Tools | Read Grep Glob Bash |
| Author | navjyotnishant |
| Source | agents/diagram-architect.md |
Returns content — does not write files
Like all but two of the agents, this one returns its output to the skill that spawned it, and the skill writes the file. It holds no write tools, so it cannot modify the repo even if asked to.
When it runs¶
The arch-diagram skill has ingested the project docs and the user wants a system architecture diagram.
generate the system architecture diagram
The arch-diagram skill spawns this agent with the repo model + type + format; the agent returns the diagram model and the rendered artifact, which the skill writes into docs/.
What it returns¶
Return the diagram model (as a short structured description) + the rendered artifact(s) + a one-line placement suggestion. For any SVG/raster output, confirm you rendered and visually verified it (Phase 3) — not just that the source is valid. Do not write files. Do not invent elements; if the repo lacks the detail for the requested type, say so and produce the best grounded diagram possible.
Spawned by¶
Derived, not declared
No agent file records which skills call it — this list is recovered from the skill definitions at build time, so it cannot go stale.
See the /tech-blog pipeline for where this fits in the whole run.