Agent Integration Test Scenarios

Crate-Researcher Tests

Test 1: Popular Crate Query

Prompt: "What's the latest version of tokio?" Expected Agent: crate-researcher Expected Sources (priority order):

1. cache/crates/tokio.json (if exists and fresh)
2. actionbook MCP → lib.rs
3. agent-browser → lib.rs
4. cargo search (last resort)

Verification:

• Agent launched correctly
• Returns version number
• Returns features list
• Cache updated after fetch

Test 2: Less Popular Crate

Prompt: "Info about the 'thirtyfour' crate" Expected Agent: crate-researcher Verification:

• Falls back correctly if not in cache
• Returns accurate info

Test 3: Cache Hit

Setup: Create cache/crates/serde.json with recent timestamp Prompt: "serde latest version" Verification:

• Returns cached data
• Response indicates "Cached: yes"
• No network fetch

---

Rust-Changelog Tests

Test 4: Specific Version Query

Prompt: "What's new in Rust 1.75?" Expected Agent: rust-changelog Expected Sources (priority order):

1. cache/rust-versions/1.75.json
2. actionbook → releases.rs
3. agent-browser → releases.rs

Verification:

• Agent launched correctly
• Returns release date
• Returns key features
• Returns stabilized APIs

Test 5: Latest Version Query

Prompt: "Latest Rust version features" Expected Agent: rust-changelog Verification:

• Determines latest version
• Returns current stable info

---

Docs-Researcher Tests

Test 6: API Documentation Query

Prompt: "How to use tokio::spawn?" Expected Agent: docs-researcher Verification:

• Fetches from docs.rs
• Returns function signature
• Returns examples
• Returns parameters

Test 7: Module Documentation

Prompt: "What's in tokio::sync?" Expected Agent: docs-researcher Verification:

• Lists module contents
• Brief descriptions

---

Clippy-Researcher Tests

Test 8: Lint Query

Prompt: "/guideline --clippy needless_clone" Expected Agent: clippy-researcher Verification:

• Returns lint description
• Maps to guideline rule
• Provides fix suggestion

Test 9: Unknown Lint

Prompt: "/guideline --clippy nonexistent_lint" Verification:

• Graceful error handling
• Suggests similar lints if possible

---

Cache Behavior Tests

Test 10: Cache Expiry

Setup:

1. Create cache/crates/test.json with timestamp 48 hours ago
2. Set TTL to 24 hours

Prompt: "test crate info" Verification:

• Detects expired cache
• Fetches fresh data
• Updates cache

Test 11: Stale-While-Revalidate

Setup:

1. Create expired cache
2. Simulate network failure

Verification:

• Returns stale data with warning
• Indicates data may be outdated

---

Error Handling Tests

Test 12: Network Failure

Setup: Simulate actionbook/agent-browser unavailable Prompt: "latest serde version" Verification:

• Falls back to cargo search
• Returns data (possibly less detailed)
• Logs the fallback

Test 13: Invalid Crate

Prompt: "info about nonexistent-crate-xyz" Verification:

• Returns "crate not found"
• Does not cache error
• Suggests similar crates if possible

---

Concurrent Agent Tests

Test 14: Parallel Crate Queries

Prompt: "Compare tokio vs async-std" Verification:

• Launches multiple agents if needed
• Aggregates results
• No race conditions in cache

Test 15: Agent + Skill Combination

Prompt: "How to use async/await with tokio?" Verification:

• m07-concurrency skill content
• tokio-specific info from agent
• Combined, coherent response

---

Performance Tests

Test 16: Cache Speed

Setup: Warm cache Prompt: "serde version" Verification:

• Response < 1 second
• No network calls

Test 17: Cold Start

Prompt: New crate query (no cache) Verification:

• Agent launches correctly
• Reasonable response time
• Cache populated for next query

你可能感兴趣的文章

Domain Skills Test Scenarios

Rust-Router Test Scenarios

Ownership (m01) Test Scenarios

Unsafe-Checker Test Scenarios

Layer 2 Skills Test Scenarios