anthropics/search-strategy

Search Strategy

If you see unfamiliar placeholders or need to check which tools are connected, see CONNECTORS.md.

The core intelligence behind enterprise search. Transforms a single natural language question into parallel, source-specific searches and produces ranked, deduplicated results.

The Goal

Turn this:

"What did we decide about the API migration timeline?"

Into targeted searches across every connected source:

~~chat:  "API migration timeline decision" (semantic) + "API migration" in:#engineering after:2025-01-01
~~knowledge base: semantic search "API migration timeline decision"
~~project tracker:  text search "API migration" in relevant workspace

Then synthesize the results into a single coherent answer.

Query Decomposition

Step 1: Identify Query Type

Classify the user's question to determine search strategy:

| Query Type | Example | Strategy | |-----------|---------|----------| | Decision | "What did we decide about X?" | Prioritize conversations (~~chat, email), look for conclusion signals | | Status | "What's the status of Project Y?" | Prioritize recent activity, task trackers, status updates | | Document | "Where's the spec for Z?" | Prioritize Drive, wiki, shared docs | | Person | "Who's working on X?" | Search task assignments, message authors, doc collaborators | | Factual | "What's our policy on X?" | Prioritize wiki, official docs, then confirmatory conversations | | Temporal | "When did X happen?" | Search with broad date range, look for timestamps | | Exploratory | "What do we know about X?" | Broad search across all sources, synthesize |

Step 2: Extract Search Components

From the query, extract:

• Keywords: Core terms that must appear in results
• Entities: People, projects, teams, tools (use memory system if available)
• Intent signals: Decision words, status words, temporal markers
• Constraints: Time ranges, source hints, author filters
• Negations: Things to exclude

Step 3: Generate Sub-Queries Per Source

For each available source, create one or more targeted queries:

Prefer semantic search for:

• Conceptual questions ("What do we think about...")
• Questions where exact keywords are unknown
• Exploratory queries

Prefer keyword search for:

• Known terms, project names, acronyms
• Exact phrases the user quoted
• Filter-heavy queries (from:, in:, after:)

Generate multiple query variants when the topic might be referred to differently:

User: "Kubernetes setup"
Queries: "Kubernetes", "k8s", "cluster", "container orchestration"

Source-Specific Query Translation

~~chat

Semantic search (natural language questions):

query: "What is the status of project aurora?"

Keyword search:

query: "project aurora status update"
query: "aurora in:#engineering after:2025-01-15"
query: "from:<@UserID> aurora"

Filter mapping: | Enterprise filter | ~~chat syntax | |------------------|--------------| | from:sarah | from:sarah or from:<@USERID> | | in:engineering | in:engineering | | after:2025-01-01 | after:2025-01-01 | | before:2025-02-01 | before:2025-02-01 | | type:thread | is:thread | | type:file | has:file |

~~knowledge base (Wiki)

Semantic search — Use for conceptual queries:

descriptive_query: "API migration timeline and decision rationale"

Keyword search — Use for exact terms:

query: "API migration"
query: "\"API migration timeline\""  (exact phrase)

~~project tracker

Task search:

text: "API migration"
workspace: [workspace_id]
completed: false  (for status queries)
assignee_any: "me"  (for "my tasks" queries)

Filter mapping: | Enterprise filter | ~~project tracker parameter | |------------------|----------------| | from:sarah | assignee_any or created_by_any | | after:2025-01-01 | modified_on_after: "2025-01-01" | | type:milestone | resource_subtype: "milestone" |

Result Ranking

Relevance Scoring

Score each result on these factors (weighted by query type):

| Factor | Weight (Decision) | Weight (Status) | Weight (Document) | Weight (Factual) | |--------|-------------------|------------------|--------------------|-------------------| | Keyword match | 0.3 | 0.2 | 0.4 | 0.3 | | Freshness | 0.3 | 0.4 | 0.2 | 0.1 | | Authority | 0.2 | 0.1 | 0.3 | 0.4 | | Completeness | 0.2 | 0.3 | 0.1 | 0.2 |

Authority Hierarchy

Depends on query type:

For factual/policy questions:

Wiki/Official docs > Shared documents > Email announcements > Chat messages

For "what happened" / decision questions:

Meeting notes > Thread conclusions > Email confirmations > Chat messages

For status questions:

Task tracker > Recent chat > Status docs > Email updates

Handling Ambiguity

When a query is ambiguous, prefer asking one focused clarifying question over guessing:

Ambiguous: "search for the migration"
→ "I found references to a few migrations. Are you looking for:
   1. The database migration (Project Phoenix)
   2. The cloud migration (AWS → GCP)
   3. The email migration (Exchange → O365)"

Only ask for clarification when:

• There are genuinely distinct interpretations that would produce very different results
• The ambiguity would significantly affect which sources to search

Do NOT ask for clarification when:

• The query is clear enough to produce useful results
• Minor ambiguity can be resolved by returning results from multiple interpretations

Fallback Strategies

When a source is unavailable or returns no results:

1. Source unavailable: Skip it, search remaining sources, note the gap
2. No results from a source: Try broader query terms, remove date filters, try alternate keywords
3. All sources return nothing: Suggest query modifications to the user
4. Rate limited: Note the limitation, return results from other sources, suggest retrying later

Query Broadening

If initial queries return too few results:

Original: "PostgreSQL migration Q2 timeline decision"
Broader:  "PostgreSQL migration"
Broader:  "database migration"
Broadest: "migration"

Remove constraints in this order:

1. Date filters (search all time)
2. Source/location filters
3. Less important keywords
4. Keep only core entity/topic terms

Parallel Execution

Always execute searches across sources in parallel, never sequentially. The total search time should be roughly equal to the slowest single source, not the sum of all sources.

[User query]
     ↓ decompose
[~~chat query] [~~email query] [~~cloud storage query] [Wiki query] [~~project tracker query]
     ↓            ↓            ↓              ↓            ↓
  (parallel execution)
     ↓
[Merge + Rank + Deduplicate]
     ↓
[Synthesized answer]