Select fingerprint guidance by applicability - #233
Merged
Conversation
nahiyankhan
marked this pull request as ready for review
July 15, 2026 14:53
chailandau
approved these changes
Jul 15, 2026
Merged
nahiyankhan
added a commit
that referenced
this pull request
Jul 16, 2026
* feat/byoa-making-loop: feat: complete the visual making loop fix: serializeNode round-trips full frontmatter (materials + free-form keys) feat: select fingerprint guidance by applicability (#233)
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Category: improvement
User Impact: Agents select brand guidance by whether it applies to the task instead of relying on vague relevance or arbitrary pull counts.
Problem:
ghost gatherlooked task-filtered even though it emitted the complete catalog, and its weak instruction left selection, pull behavior, and fingerprint silence underspecified.Solution: Make the complete, unfiltered, unranked contract explicit; select every applicable node and no inapplicable node; direct agents through
ghost pull; expose concrete payload types.Validation:
pnpm test: 175 passed, 1 skippedpnpm exec vitest run packages/ghost/test/cli.test.ts packages/ghost/test/ghost-core/catalog-assemble.test.ts: 69 passed, 1 skipped after rebasing onto currentmainpnpm --filter @design-intelligence/ghost build: passedpnpm typecheck: passedpnpm exec biome check .: passedpnpm check:terminology: passedpnpm check:cli-manifest: passedpnpm check:release-workflows: passedpnpm check: blocked only atcheck:packed-packageby repeated artifact-registryECONNRESET; all remaining local checks passed individuallyRelease (next)smoke publish:@design-intelligence/ghost@0.0.0-next-20260715121453Changeset: added as a minor change.
Ghost Review:
ghost check --base main: not run because this CLI has nocheckcommandghost review --base main --include-memory: not run because--include-memoryis not supported by the current review commandFile changes
.changeset/gather-applicability-fences.md
Records the public gather-contract and fence-safety change.
CLAUDE.md
Aligns repository agent guidance with applicability-based gather selection.
apps/docs/.ghost/pattern.docs-index-card-grid.md
Updates the docs fingerprint example away from the retired gather phrasing.
apps/docs/src/app/tools/scan/page.tsx
Updates the scan tool description to explain applicable-node selection.
apps/docs/src/content/docs/cli-reference.mdx
Documents the complete menu, applicability contract, explicit pull step, and silence handling.
apps/docs/src/content/docs/fingerprint-authoring.mdx
Teaches descriptions as condition-bearing retrieval payloads.
apps/docs/src/content/docs/getting-started.mdx
Updates the gather-to-pull workflow and artifact-versus-menu terminology.
apps/docs/src/content/docs/troubleshooting.mdx
Explains how to diagnose applicability and description failures.
apps/docs/src/generated/cli-manifest.json
Regenerates the public CLI manifest for the updated gather help text.
docs/purposes.md
Aligns architectural purpose text with applicability-based selection.
packages/context-control/lib/model.mjs
Removes count-based selection doctrine from the context-control selector prompt.
packages/ghost/src/commands/command-discovery.ts
Updates gather command discovery copy.
packages/ghost/src/commands/gather-command.ts
Emits the markdown and JSON applicability contract, cover status, silence behavior, and payload breakdown.
packages/ghost/src/commands/pull-command.ts
Uses dynamically sized Markdown fences for inlined materials and Skeletons.
packages/ghost/src/ghost-core/catalog/assemble.ts
Derives substantial fenced-example metadata separately from Skeleton content.
packages/ghost/src/ghost-core/catalog/menu.ts
Adds fenced-example metadata to menu entries.
packages/ghost/src/ghost-core/catalog/types.ts
Carries the additive payload metadata in catalog node types.
packages/ghost/src/ghost-core/index.ts
Exports the substantial fenced-example helper.
packages/ghost/src/ghost-core/node/index.ts
Exports the node-level helper through the node module.
packages/ghost/src/ghost-core/node/steering.ts
Defines substantial fenced examples as fences outside Skeleton sections.
packages/ghost/src/ghost-core/node/types.ts
Aligns node frontmatter typing with preserved passthrough metadata.
packages/ghost/src/init-payloads/skeleton/brand.md
Distinguishes cover delivery status from universal applicability.
packages/ghost/src/scan/fingerprint-package-lint.ts
Aligns description diagnostics with applicability-based retrieval.
packages/ghost/src/skill-bundle/SKILL.md
Updates the core skill contract for gather, pull, and silence.
packages/ghost/src/skill-bundle/references/blocks.md
Clarifies how concrete payloads support applicable guidance.
packages/ghost/src/skill-bundle/references/brief.md
Removes arbitrary pull-count guidance and briefs from applicable nodes instead.
packages/ghost/src/skill-bundle/references/capture.md
Teaches authors to state what a node governs and when it applies.
packages/ghost/src/skill-bundle/references/recall.md
Updates recall to select every applicable node and use
ghost pull.packages/ghost/src/skill-bundle/references/schema.md
Documents the additive menu metadata and selection contract.
packages/ghost/src/skill-bundle/references/steering-audit.md
Aligns audit language with gather payload labels and existing pulse output.
packages/ghost/test/cli.test.ts
Covers gather contracts, payload labels, cover state, silence, and nested fence safety.
packages/ghost/test/ghost-core/catalog-assemble.test.ts
Verifies Skeleton-only nodes are not double-classified as fenced examples.
Screenshots/Demos: N/A; this changes CLI and machine-readable output rather than a visual surface.