Scout search modes for coding agents

Search guide

Scout search modes

Choose a search mode from the question. Exact scans and ranked discovery answer different questions.

Use it for exact, exhaustive lookup.

literal finds an exact text sequence.
regex finds text that matches a regular expression.
word finds whole words and excludes substring-only matches.
Strict results report whether the scan was exhaustive. Partial scans report their limit and coverage.

{ "query": "deprecated_api", "target": "content", "match": "word" }

Use it for approximate file, path, or identifier discovery.

{ "query": "usrctrl", "target": "files", "match": "fuzzy" }

Use it for concept discovery when the query and code use different words.

Semantic queries find related code concepts. The optional tier uses a local model.
If the tier is unavailable, degraded reports a lexical fallback.

{ "query": "where request retries are limited", "target": "semantic" }

Use it when concept similarity and shared vocabulary both matter.

Semantic queries fuse embedding search with a lexical ranking over the same chunks.
route.tier reports hybrid when both rankings contribute.
The returned score orders that result set. It is not a similarity percentage.

Use it for syntax-aware matching by code shape.

target: structural runs ast-grep patterns with $NAME and $$$ captures.
target: symbols returns a declaration outline with names, kinds, and containers.
Native syntax support depends on a linked grammar. Scout reports degradation for uncovered languages.

{ "query": "console.log($$$)", "target": "structural", "scope": { "languages": ["javascript"] } }

Use the same search surface across repeated agent steps.

Scout keeps a resident workspace index and updates it from filesystem events.
Each query applies pending watcher changes before it reads the index.
server_status reports index revision, file count, tier state, and watcher state.

Scope searches with path prefixes, globs, or language filters. Check complete, reason, coverage, and degraded before accepting an empty result.
Semantic search is optional. Read its requirements and setup.
Use Glider or TGlider for compiler-backed references and refactors.
See benchmarks for measured latency, setup costs, conditions, and unmeasured areas.