Menu
Use when ingesting, refreshing, listing, searching, or deleting a local corpus with Blackbird code search and `gh blackbird search --fileset`.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Use when reaching for `gh blackbird` (Blackbird code search) for cross-repo lexical, symbol, or semantic search on GitHub — finding callers, ownership, or how systems work without cloning.
Use when creating a GitHub pull request, or when updating an existing PR's title or body so it matches what the code actually does.
Use when authoring or substantially editing a design doc, architecture doc, or subsystem explanation — the "here's what's there and why" companion to an ADR's terse decision record.
Use when creating a repo-tracked multi-agent planning PR for a large project, especially when phases, living docs, parallel agent prompts, and cross-PR coordination are needed.
Use when about to call a library, crate, or framework API you haven't verified, when a dependency's behavior is surprising, when training-data memory might be stale, or when investigating how a dependency actually behaves.
Use when investigating an alert, monitor, incident, error spike, degraded service, or production anomaly, and the repository has no incident-investigation skill of its own.
| name | blackbird-fileset |
| description | Use when ingesting, refreshing, listing, searching, or deleting a local corpus with Blackbird code search and `gh blackbird search --fileset`. |
Use external filesets for code-like corpora that are on disk but are not indexed as GitHub repositories. This skill owns the fileset lifecycle; use blackbird for general GitHub code-search strategy.
From the corpus directory, keep ingestion and search on the same inferred name:
gh blackbird fileset ingest --dry-run
gh blackbird fileset ingest
gh blackbird search --fileset --semantic "where is request throttling implemented?"
gh blackbird search --fileset 'rate_limit'
Bare fileset ingest recursively ingests the current directory and names the fileset after its canonical basename. Bare search --fileset independently infers the current directory basename. Run both from the same directory, or pass an explicit pair:
gh blackbird fileset ingest ./vendor/widget --name widget-src
gh blackbird search --fileset widget-src --semantic "how are retries bounded?"
--semantic for a concept. External fileset search rejects --symbol, --auto-index, and --experiment.--max-tokens N overrides that lexical budget and has no effect on semantic search.--format oneline falls back to pretty. -A/-B/-C/-M/--full-snippet do not shape external results and disable the implicit token budget.--dry-run first for an unfamiliar or large tree. It walks, filters, hashes, and reports the checkpoint without making HTTP calls..gitignore and skips target, node_modules, .git, data, and .jj. Linguist filters binary, generated, and vendored content. Non-UTF-8 paths are skipped, and one fileset is capped at 200,000 ingestable files.~/.config/gh/blackbird/. Re-run the same command to resume. Use --restart only when the stored ingest expired or the server aborted it; server ingest state lasts about two hours.search_hint but does not pass it to search. Retry after a few seconds instead of re-ingesting.Filesets belong to the authenticated GitHub actor:
gh blackbird fileset list --json
gh blackbird fileset delete widget-src
gh blackbird fileset delete widget-src --yes
List before diagnosing a 404 or deleting by name. Deletion is irreversible; --yes is required when stdin is not interactive.