khujta Gabe Roast — Adversarial Gap Review Skill
Gabe execution contract (E1–E7)
This skill runs under the suite execution contract — E1 EVIDENCE · E2 RUN-BEFORE-✅ · E3 NO SILENT DOWNGRADE · E4 REUSE FIRST · E5 STATE SYNC · E6 MISSING ANCHOR = STOP · E7 REPORT WHERE — floors, not ceilings; a skill's own gate may be stricter, never looser. Full text: ../gabe-docs/references/execution-contract.md (if that file is missing, E6 applies — STOP).
What this does
Stress-tests a target (file, folder, plan, architecture, or concept) from a required perspective to surface gaps, risks, and missing pieces — classified by maturity level (MVP/Enterprise/Scale) and importance (Critical/High/Medium/Low), with actionable output that drives decisions. Not a generic checklist: the skill fully adopts the named perspective, reads deeply, and attacks like someone whose job depends on finding what's wrong.
Usage / modes
/gabe-roast [perspective] [target]
Two inputs are required — target (file, folder, inline plan/concept, or "this conversation") and perspective (never optional; examples: Architect, UX/UI Designer, Security Auditor, QA/Testing Lead, DevOps Engineer, Domain Expert, End User — or a canonical work-mode archetype: Prototyper / Builder / Sweeper / Grower / Maintainer, defined in templates/archetype-map.md). If either is missing, ask before proceeding.
| Mode | Output |
|---|---|
| full (default) | Per-gap fields (Gap, One-liner, Effort, Lose, Evidence, optional Fix), grouped by maturity level then importance |
| brief | Scannable table: ID / Maturity / Importance / Gap / One-liner / Effort |
Procedure
- Treat any text after the invocation as
$ARGUMENTS(perspective + target). - Read
references/roast-spec.mdIN FULL before executing — the binding spec. If missing, E6 applies — STOP. - Confirm both target and perspective are present; ask if either is missing.
- Pre-roast gate: run
/gabe-assess briefon the target (skip if no.kdbp/VALUES.mdor~/.kdbp/VALUES.mdexists). PASS → proceed; CONCERN → warn and proceed; FAIL → warn and confirm before continuing. - Read before attacking: list the folder tree, open every entry point + config, never cite a file not opened this session, print a read ledger (
Read: 14/38 files — skipped: ...) directly under the perspective line. - Stay fully in character for the perspective; classify each gap by maturity level and importance, filling every required per-gap field with cited evidence (or a recorded 0-hit search for absence claims).
- Kill-gate (mandatory before printing): re-verify every gap against its Evidence; a gap with an empty Evidence field is deleted, not demoted. Print
drafted N → killed X → reported Yabove the total line. - Render the requested mode's output; if zero gaps survive, say so explicitly rather than manufacturing findings.
Output contract (summary)
Full mode: GABE ROAST header (target, perspective, read ledger) + gaps grouped by maturity (MVP/Enterprise/Scale) then importance (Critical/High/Medium/Low), each with Gap/One-liner/Effort/Lose/Evidence/Fix, ending in a mandatory TOTAL: [X] gaps — ... summary line plus the kill-gate stats. Brief mode: the same gaps as one table row each, sorted maturity-then-importance-then-impact, no Evidence/Lose/Fix columns. The full output contract in the spec is binding.