refactor: extract scorePrompt into lib + 20 tests + scope scoring fix#290
Open
TerminalGravity wants to merge 6 commits intomainfrom
Open
refactor: extract scorePrompt into lib + 20 tests + scope scoring fix#290TerminalGravity wants to merge 6 commits intomainfrom
TerminalGravity wants to merge 6 commits intomainfrom
Conversation
Adds a ready-to-use CLAUDE.md template that makes Claude Code automatically run preflight_check on prompts. Users can copy it into their project to get preflight working without manual tool calls. Referenced from Quick Start in README and examples/README.
- CLI now responds to --help/-h with usage info, profiles, and links - CLI now responds to --version/-v with package version - Previously, any flag just launched the interactive wizard - Fixed README badge from Node 18+ to Node 20+ (matches engines field)
Adds a new 'export_timeline' MCP tool that generates markdown reports from timeline data with: - Summary stats table (events by type, correction rate, commits/prompt) - ASCII activity chart grouped by day - Recent commits log - Correction insights - Error summary - Configurable period (day/week/month) with offset - Optional save to ~/.preflight/reports/ Includes tests (2 passing). Closes #5
Adds 10 tests covering: - Trivial prompt pass-through - force_level=skip bypass - Ambiguous prompt detection (vague pronouns, short prompts, vague verbs) - Multi-step execution plan generation - Git state inclusion in clarification - Triage confidence/reasons display - Pattern match triage boosting - Risk level assignment in execution plans Brings test count from 45 to 55.
Several tools were passing shell syntax (pipes, redirects, || chains) and non-git commands (cat, find, pnpm, gh) to run(), which uses execFileSync without a shell. This caused silent failures: - Shell redirections (2>/dev/null) passed as literal git args - Pipe chains (| grep, | tail) passed as literal git args - 'git' prefix doubled (run already prepends 'git') - Non-git commands (find, gh, pnpm) routed through git execFileSync Fix: - Add shell() helper using execSync for commands needing shell features - Harden run() string parsing to strip leading 'git' and shell syntax - Migrate 10 call sites across 8 tools to use shell() or proper arrays - Add 13 tests covering run() cleanup and shell() behavior Affected tools: audit-workspace, clarify-intent, enrich-agent-task, sequence-tasks, session-handoff, session-health, verify-completion
…ng bug - Extract pure scorePrompt() into src/lib/prompt-scoring.ts for testability - Add comprehensive test suite (20 tests) covering all 4 dimensions + grading + edge cases - Fix scope scoring: long prompts alone no longer get max score (25→20), require explicit bounding keywords for full marks - Tool now imports from shared lib instead of inlining logic
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
scorePrompt()pure function fromprompt-score.tsintosrc/lib/prompt-scoring.tsWhy
The scoring logic was inlined in the tool registration, making it untestable. Now it's a pure function with full coverage.
Test results
88 tests passing (20 new + 68 existing)