Files
LexAI/docs/HANDOFF.md
john kevin asprec 444060c3eb Add new agents and skills for enhanced project orchestration and review processes
- Introduced `critic`, an independent adversarial reviewer for security and correctness.
- Added `fable-orchestrator` to manage task routing and verification.
- Implemented `gauntlet-critic` for fresh-context evaluation of gauntlet rounds.
- Created `planner` for generating executable implementation plans with dependencies.
- Developed `security-auditor` for application security reviews and audits.
- Established `system-steward` to improve agent prompts and skills based on verified failures.
- Added `dev-loop` skill for autonomous development loops over repositories.
- Implemented `gauntlet-loop` skill for iterative quality benchmarking against reference standards.
- Updated project settings to utilize the new orchestrator agent.
- Created documentation for `GAUNTLET.md`, `PROGRESS.md`, and `REFERENCE_BAR.md` to track project status and quality benchmarks.
- Added detailed prompting style guide to enhance understanding of prompt patterns and agentic loops.
2026-08-08 16:49:07 +08:00

17 lines
1.3 KiB
Markdown

# Handoff — LexAI
## Handoff — 2026-08-08
Outcome: done — Prompt Builder pattern work verified; in-page live hint added.
Delivered: `src/lib/actions.ts` — 12-entry `PROMPT_PATTERNS` (Direct/Reasoning/Agentic groups), "style" now "pattern".
Delivered: `src/lib/providers.ts` `getSystemPrompt` — ROLE+TASK → pattern block (auto rubric or one pattern's skeleton+guard) → persona/format/style modifiers → invariants last.
Delivered: `entrypoints/background.ts:71``prompt` action gets a 2048-token `maxTokens` floor (was truncating under the 1024 input-length floor).
Delivered: `src/lib/actions.ts` `resolvePromptPattern` — migrates legacy `promptStyle` storage values.
Delivered: `entrypoints/content.ts:722-733` — in-page dialog now shows the live per-pattern hint (popup already had it); updates on select `change` and after storage prefill.
Verified: `npm run typecheck` — clean.
Verified: `npm test -- --run` — 64 tests, 4 files, all passing.
Verified: `npm run build``.output/chrome-mv3/` built, no errors.
Decisions: none new this session.
Known risks: real-page load-unpacked check still outstanding — this is the owner's gate (see `docs/PROGRESS.md`).
Next smallest action: load `.output/chrome-mv3` unpacked and confirm the Pattern dropdown + hint render correctly in both popup and in-page dialog, and a saved pattern preselects.