| name | resume-debate |
| description | Structured 3-round debate about a single resume bullet. Runs hiring-manager, bar-raiser, and future-self-skeptic through opening critique, rebuttal, and rewrite rounds. Produces either a consensus rewrite or an articulated disagreement for the user to adjudicate. Accepts either the bullet text or a line number as $ARGUMENTS. Use when the user is uncertain about a specific bullet, asks "is this bullet good?", wants multiple perspectives on one line, or says a reviewer flagged a specific bullet. |
| argument-hint | <bullet-text-or-line-number> |
| allowed-tools | Read, Grep, Glob, Agent, AskUserQuestion |
Bullet Debate
Orchestrate a 3-round structured debate between three personas about a single resume bullet, then synthesize the result.
Process
-
Resolve the bullet. If $ARGUMENTS is a line number (e.g., 34, line 34), read will_cygan_resume.typ and extract that line. If it's a full bullet text, use it directly. If $ARGUMENTS is empty, use AskUserQuestion to prompt for either a line number or the bullet text.
-
Run the 3 rounds. Each round is a single parallel Agent tool message with 3 calls (one per persona). See REFERENCE.md § Round Protocol for the exact prompts to send at each round. Rounds are sequential — round 2 depends on round 1's output, round 3 depends on rounds 1 and 2.
-
Synthesize. Compare the 3 candidate rewrites from round 3. Apply the synthesis rules in REFERENCE.md § Synthesis Rules.
-
Report. Use the output format in REFERENCE.md § Output Format.
Agent teams note
If the user wants the three personas to truly argue in real time via SendMessage (rather than sequential rounds), that requires setting CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS and using TeamCreate. The default here is the sequential sub-agent approach — it works today, is deterministic, and produces the same debate structure.
Notes
- Full disagreement is a valid outcome. If the three personas cannot converge after round 3, present all three rewrites side-by-side and let the user decide. Do not force consensus.
- Keep each round's prompts terse. Long prompts lead to long sub-agent responses that dilute the debate.
See REFERENCE.md for the round protocol, synthesis rules, and output template.
