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.
npx skills add luuuc/sense --skill smitherygit clone --depth 1 https://github.com/luuuc/senseWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/luuuc/sense/smithery)<a href="https://agentmods.dev/skills/luuuc/sense/smithery"><img src="https://agentmods.dev/badge/skills/luuuc/sense/smithery/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/luuuc/sense/smithery"><img src="https://agentmods.dev/badge/skills/luuuc/sense/smithery.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00054 | $0.01060 |
| Opus 5.5 | $0.00022 | $0.00424 |
| Sonnet 5.5 | $0.00011 | $0.00212 |
| Haiku 4.5 | $0.00005 | $0.00106 |
Grade C, and why
sense scanned grade C with 2 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 1mo 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.
Downloads and executes remote codehighSupply chain
curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.
curl -fsSL https://luuuc.github.io/sense/install.sh | sh Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -fsSL https://luuuc.github.io/sense/install.sh | sh How it starts
The opening of the file, as written. The whole thing — 92 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Sense: Codebase Understanding for AI Coding Agents
Sense is an MCP server that gives AI agents the structural understanding of a codebase that a senior engineer carries in their head. It runs as one Go binary with a local SQLite index — no API keys, no SaaS, no cloud calls.
When to Use This Skill
Use Sense whenever the task depends on understanding code relationships, not just reading text:
- "Who calls this function?" / "What does this function call?"
- "What breaks if I change this signature?" (blast radius)
- "Find the implementation of the X pattern" (semantic search)
- "What conventions does this project follow?" (e.g., naming, Rails idioms, Go interface patterns)
- "Find dead code"
- "Show me all routes / handlers / models / controllers"
Do NOT use Sense for:
- Locating a string literal or log message — plain grep is faster.
- Editing code — Sense is read-only by design.
Setup (one-time per project)
# Install the binary (macOS / Linux):
curl -fsSL https://luuuc.github.io/sense/install.sh | sh
# In the project root:
sense scan # builds the .sense/ index (tree-sitter + embeddings)
sense setup # writes .mcp.json + CLAUDE.md routing + hooks
After sense setup, your AI tool (Claude Code, Cursor, Codex CLI) has Sense's four tools available via MCP.
Tools
| Tool | Capability |
|---|---|
sense_graph |
Symbol relationships — callers, callees, inheritance, tests, dead code |
sense_search |
Hybrid semantic + keyword search with text fallback |
sense_blast |
Blast radius, affected code, affected tests, risk score |
sense_conventions |
Auto-detected project conventions (Rails, React, Django, Go, naming) |
Plus a generated .sense/summary.md at session-start: top namespaces, hub symbols, entry points, conventions — so the AI is oriented before its first tool call.
How to Call (MCP)
After sense setup, the tools are exposed automatically. Prefer them over grep/glob for any structural question:
- "Who calls X?" →
sense_graph symbol="X" direction=callers - "What would break if I change Y?" →
sense_blast symbol="Y" - "Find code that does Z" →
sense_search query="natural language description" - "How does this project define services?" →
sense_conventions
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.
- 1mo ago First seen · 92 lines · 54 tokens per session scan C f358615a0a53
sense is a skill published in the GitHub repository luuuc/sense (40 stars, last pushed yesterday), licensed MIT. It adds 54 tokens to every session and 1,060 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
nexus-query
Precise, instant code structure queries for active development — answer 'who depends on this interface before I refactor it', 'how many modules break if I change this', 'what is the real impact radius of this feature change', 'which module is the true high-coupling hotspot in this legacy codebase'. Essential before…
omh-live-incident-response
This is an OMH live-incident-response workflow skill, projected for Agent Skills hosts (Claude Code, Codex, Cursor, opencode, OpenClaw, pi).
ix
Answer structural questions about a codebase — what a symbol is, what calls it, what a change breaks, where the hotspots are — from a persistent code graph via the ix CLI, instead of grepping.
report-issue-local
File a bug or feature request against this MCP server's own repo. Use for server-specific issues — tool logic, service integrations, config problems, or domain bugs that aren't caused by the framework.
debugging
Use this skill when something is broken and the cause is unknown — a failing test, a crash, an error message, wrong output, a regression, a hang, or behaviour that differs between environments — and it has to be found and fixed at the root. Also use it to build the failing proof for a suspected defect before fixing…
unicli-repair
Evidence-driven repair workflow for a broken Uni-CLI adapter. Trigger on a failed unicli envelope, a quarantined adapter, or an explicit adapter-repair request. Classifies non-source failures, edits only the reported adapter path, and uses the original command as a bounded oracle.