> For the complete documentation index, see [llms.txt](https://docs.heeler.com/mrecEO40m5D6bt7Pq5pE/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.heeler.com/mrecEO40m5D6bt7Pq5pE/catalog/files.md).

# Agent and Instruction Files

An inventory of notable repository files — above all AI-agent instruction files like CLAUDE.md — each scored for risk, recording what your coding agents are instructed to do.

The **Files** inventory (**Catalog → Files**) catalogs notable repository files that warrant security scrutiny — most importantly **AI-agent instruction files** such as `CLAUDE.md`, `AGENTS.md`, Cursor rules, and skill definitions — across every major coding assistant, not just Claude — and scores each one for risk. As coding agents increasingly read these files to decide how to write, modify, and remediate code, a manipulated instruction file becomes a real supply-chain attack vector. Heeler analyzes them directly from source and reports what each file instructs an agent to do.

<figure><img src="https://414480750-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FXP3dp2kecwKA2KvYkntz%2Fuploads%2Fgit-blob-8c6352bc4bad6ab273ff01970b74750c34a8e9ac%2Fcc-cat-files.png?alt=media" alt="The Files inventory with kind, score, assessed intent, and findings."><figcaption><p>Files — instruction files, hooks, skills, and configs, each with a risk score and an assessed intent of Benign, Suspicious, or Malicious.</p></figcaption></figure>

{% hint style="info" %}
**Prerequisites:** The Files inventory is built from source analysis of your connected source-code (SCM) provider, wherever the Agent Files preview is enabled.
{% endhint %}

## The listing

| Column              | What it shows                                                                          |
| ------------------- | -------------------------------------------------------------------------------------- |
| **Repository**      | The repo and SCM organization the file belongs to.                                     |
| **Path**            | The file path, linked to source (bundles expand to show their members).                |
| **Kind**            | The file's classification — **Instructions**, **Hook Script**, **Skill**, **Config**.  |
| **Score**           | The overall analysis score (0–100), shown with an **At Risk** / **Not at Risk** state. |
| **Assessed Intent** | The verdict — **Benign**, **Suspicious**, or **Malicious**.                            |
| **Findings**        | Security findings detected in the file.                                                |
| **Last Modified**   | When the file last changed in source.                                                  |
| **Ticket**          | The tracking ticket linked to the file, once one has been opened for it.               |

Filter by Repository, Assessed Intent, Kind, and **At Risk**, and search across paths; admins can save filtered views. See [Filtering and Exports](/mrecEO40m5D6bt7Pq5pE/operate/dashboards/filtering-and-exports.md).

{% hint style="info" %}
**Lower score = higher risk.** A file scored 0 with a **Malicious** intent is the one to open first; a file at 45 marked **Benign** is not at risk. The **At Risk** filter isolates the ones that need attention.
{% endhint %}

## Reading a file

Selecting a file opens its [**detail view**](/mrecEO40m5D6bt7Pq5pE/catalog/files/file-detail.md) — the risk score broken into its static, LLM, and external components, a plain-language summary of what the file contains and why it got its verdict, and the specific findings behind it.

## Related

* [File Detail](/mrecEO40m5D6bt7Pq5pE/catalog/files/file-detail.md) — the score breakdown, summary, and findings.
* [Agent Files](/mrecEO40m5D6bt7Pq5pE/findings/agentic-supply-chain-agent-files.md) — how these files are prioritized and triaged as findings.
* [MCP Server Analysis](/mrecEO40m5D6bt7Pq5pE/catalog/endpoints/mcp-analysis.md) — the MCP tooling these instruction files often drive.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.heeler.com/mrecEO40m5D6bt7Pq5pE/catalog/files.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.