refine

A set of manual checks for RAGgrep, a search system that retrieves relevant files from a codebase. The checks test whether common searches return the expected source files or functions near the top of the results.

In plain words
What is it for?
Use it to index the example project and test searches for authentication, databases, password hashing, sessions, email, caching, and JWT token handling.
Why use it?
It provides a repeatable way to spot poor search ranking before changes to the search system are accepted.

Command for Cursor

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/conradkoh/raggrep/refine
Clone the repo
git clone --depth 1 https://github.com/conradkoh/raggrep

Made for: Cursor.

Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,029 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.01029
Opus 5 $0.00000 $0.00515
Sonnet 5 $0.00000 $0.00206
Haiku 4.5 $0.00000 $0.00103

Measured 2d ago against content hash 9169fe638af9, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

refine scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.cursor/commands/refine.md · 116 lines

How it starts

The opening of the file, as written. The whole thing — 116 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Refine RAGgrep Search Quality

Run manual sanity checks on the RAGgrep search system to ensure ranking quality.

Instructions

  1. Setup the test scenario (if not already done):

    • The scenarios/basic folder contains realistic test files
    • Run cd scenarios/basic && bun run ../../src/app/cli/main.ts index to index
  2. Run sanity checks:

    cd scenarios/basic
    
    # Test 1: User authentication (expect: src/auth/login.ts)
    bun run ../../src/app/cli/main.ts query "user authentication" --top 5
    
    # Test 2: Database connection (expect: src/database/connection.ts)
    bun run ../../src/app/cli/main.ts query "database connection pool" --top 5
    
    # Test 3: Password handling (expect: hashPassword function)
    bun run ../../src/app/cli/main.ts query "password hashing" --top 5
    
    # Test 4: Session management (expect: src/auth/session.ts)
    bun run ../../src/app/cli/main.ts query "session validation" --top 5
    
    # Test 5: Email service (expect: src/services/email.ts)
    bun run ../../src/app/cli/main.ts query "send welcome email" --top 5
    
    # Test 6: Cache (expect: src/services/cache.ts)
    bun run ../../src/app/cli/main.ts query "redis cache" --top 5
    
    # Test 7: JWT (expect: verifyToken function)
    bun run ../../src/app/cli/main.ts query "JWT token verification" --top 5
    
    # Test 8: API routes (expect: src/api/routes/users.ts)
    bun run ../../src/app/cli/main.ts query "user registration endpoint" --top 5
    
    # Test 9: JSON config files
    bun run ../../src/app/cli/main.ts query "rate limit configuration" --top 5
    
    # Test 10: SQL files
    bun run ../../src/app/cli/main.ts query "seed test users" --top 5
    
  3. Analyze ranking results:

    • Top result should match expected file
    • Score should be > 40% for good matches
    • Related files should appear in top 5
  4. If ranking issues found:

    • Create a formal test case in src/tests/integration.test.ts
    • Investigate the scoring in src/domain/services/bm25.ts or src/modules/core/index.ts
    • Check tokenization in src/domain/services/keywords.ts
    • For semantic search, check src/modules/language/typescript/index.ts

Read the full file on GitHub · 116 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 116 lines · 0 tokens per session scan A 9169fe638af9

Subscribe to this mod's changes

refine is a command published in the GitHub repository conradkoh/raggrep (22 stars, last pushed 4mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,029 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.