spec

/aide:spec — Spec Phase

Agent: This command is executed by the aide-spec-writer agent.

Produce the .aide intent spec frontmatter only. This is the spec-writing phase — the session that distills the orchestrator's delegation context into a falsifiable intent contract. The orchestrator owns the user conversation and passes the gathered context in the delegation prompt. Body sections (Context, Strategy, Good/Bad examples, References) are conditional — they exist only when the orchestrator routes through /aide:synthesize afterwards. Most modules are navigation stubs and skip synthesize entirely, going straight to /aide:plan. Signal whether synthesize is needed in your return.

Checklist

• Read the delegation context from the orchestrator. If insufficient to write specific outcomes (missing: what the module is for, who consumes its output, what success looks like, what failure looks like), return to the orchestrator listing what's missing
• Read the AIDE template before writing — copy the fenced template block from the canonical template doc into the new file
• Decide filename:
  • Use .aide if no research.aide exists in the target folder
  • Use intent.aide if research.aide exists in the same folder (co-located research is an escape hatch — prefer the brain)
• Fill the frontmatter ONLY against hard caps:
  • scope — the module path this spec governs
  • description — one sentence, ≤ 200 characters
  • intent — one paragraph, ≤ 100 words, ten-second north star
  • outcomes.desired — 3-6 items, ≤ 2 sentences each, falsifiable
  • outcomes.undesired — 3-6 items, ≤ 2 sentences each, the almost-right-but-wrong kind
• Body sections are conditional. Default: produce a frontmatter-only file with NO body section headings. If the orchestrator's delegation prompt explicitly says synthesize will run, then preserve empty body section placeholders for the strategist; otherwise omit them entirely.
• No code in the spec — no file paths, no type signatures, no function names, no argument indexes. Domain vocabulary IS allowed (entity names, status enums, field names that name a domain concept).
• Every outcomes entry must trace back to the intent paragraph. Cut any outcome that doesn't
• Present the frontmatter to the orchestrator for relay to the user
• Run aide_validate to check the spec for structural issues
• Return two routing signals to the orchestrator: Research needed: yes/no and Strategy needed: yes/no. Default both to NO unless the interview surfaces genuine domain complexity. The orchestrator decides next stage:
  • both NO → /aide:plan directly (orchestrator passes user implementation context)
  • Strategy YES, Research NO → /aide:synthesize (brain already has research)
  • both YES → /aide:research first, then /aide:synthesize, then /aide:plan