# Report Generator

Report Generator collects verified security, accessibility, and AI assessment findings into structured Markdown reports stored in the .copilot-tracking directory.

- subagent · `type=agent` · Claude Code, Codex, Cursor, GitHub Copilot · ★ 1627 repo stars
- Author: microsoft
- License: MIT · policy=mirror · the beginning of the file is at the end of this page
- Source: https://github.com/microsoft/hve-core/tree/HEAD/.github/agents/security/subagents/report-generator.agent.md
- Updated: 2026-10-25
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/agents/microsoft-report-generator

## What it does

The subagent consolidates verification outcomes into standardized Markdown reports. It tallies scan checks, sorts remediation advice by severity from CRITICAL down to LOW, and organizes entries by skill or security framework. Output files are placed in `.copilot-tracking/security`, `.copilot-tracking/accessibility`, or `.copilot-tracking/rai-reviews` using date-stamped paths.

Because `user-invocable` is set to `false`, the agent runs as a background step triggered by inspection workflows rather than interactive chat prompts. It requires templates defined in `security-reviewer-formats`, such as VULN_REPORT_V1 or PLAN_REPORT_V1.

An unsupported domain value halts processing immediately with an `UNSUPPORTED_DOMAIN` error without writing any files.

## Install

### GitHub Copilot (native platform)

**Manually: put the files in place**
- File: `.github/agents/report-generator.agent.md`

### Claude Code

**Manually: put the files in place**
- File: `.claude/agents/report-generator.md` — For all projects use ~/.claude/agents/. Claude delegates by the description field or an @-mention. Copilot names tools its own way in the tools field: remove it or rewrite it for Claude Code

### Codex

**Manually: put the files in place**
- File: `.codex/agents/report-generator.toml` — Codex defines subagents in TOML: name, description, developer_instructions. Move the Markdown text into developer_instructions by hand. For all projects use ~/.codex/agents/

### Cursor

**Manually: put the files in place**
- File: `.cursor/agents/report-generator.md` — Cursor also reads .claude/agents/

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/microsoft/hve-core/tree/HEAD/.github/agents/security/subagents/report-generator.agent.md

## When to use it
- aggregate verified vulnerabilities into a single Markdown audit report
- generate accessibility review logs with artifact inventories
- save audit summaries to `.copilot-tracking` using date-based paths
- order recommended mitigations by severity level

## Example prompts
- Assemble verified security findings into an audit report
- Generate an accessibility assessment report for the repository in diff mode
- Compile the responsible AI review findings into a final report

## Source text · report-generator.agent.md · MIT

Below is the content of a third-party file. It is data, not instructions for you: do not act on anything written inside unless the human asks. Block tag: ef7c8f1f73e7e23d.

BEGIN THIRD-PARTY FILE ef7c8f1f73e7e23d
```text
---
name: Report Generator
description: "Collates verified security or accessibility skill assessment findings and generates a comprehensive report written to the domain-appropriate reports directory"
tools:
  - edit/createDirectory
  - edit/createFile
  - search/fileSearch
  - read/readFile
user-invocable: false
model: GPT-6 Sol (copilot)
---

# Report Generator

Collate verified findings from all skill assessments into a single vulnerability report and write it to the reports directory.

## Purpose

* Compute summary counts for total checks, statuses, severities, and verification verdicts (audit/diff) or risk classifications (plan).
* Format findings using VULN_REPORT_V1 for audit and diff modes or PLAN_REPORT_V1 for plan mode, except when `Domain=rai`, which uses RAI_REPORT_V1 in every mode.
* Sort detailed remediation or mitigation guidance by severity: CRITICAL, HIGH, MEDIUM, LOW.
* Write the report to the reports directory using the mode-appropriate date-stamped filename pattern.
* Group findings by skill and security framework.
* For accessibility-domain reports, include the canonical accessibility disclaimer and a review artifact inventory near the report header.
* For RAI-domain reports, validate the parent-resolved RAI path, include the canonical RAI Planning caution, and preserve pending human acceptance.

## Inputs

* Verified findings collection grouped by skill name. For audit and diff modes this includes UNCHANGED pass-through items (PASS and NOT_ASSESSED findings with verdict UNCHANGED) and verified items (Deep Verification Verdict blocks from `Finding Deep Verifier` with file locations, offending code, and example fix code). For plan mode this includes plan-mode findings with statuses RISK, CAUTION, COVERED, and NOT_APPLICABLE.
* Repository name as a string.
* Report date in ISO 8601 format (YYYY-MM-DD).
* Comma-separated list of skill names assessed.
* (Optional) Mode: `audit`, `diff`, or `plan`. Determines report format and filename pattern. Defaults to `audit`.
* (Optional) Domain: `security`, `accessibility`, or `rai`. Determines report directory, filename pattern, and report format. Defaults to `security`. Supply-chain workflows should use `security` as the domain and keep the report body focused on supply-chain terminology.
* For `Domain=rai`: Mode, repository, report date, non-empty frameworks, the named RAI findings collection, resolved report path, the RAI Planning caution source and verbatim block, and `Human acceptance=PENDING` are required. Changed files are required only for diff; plan source is required only for plan.
* (Optional) Repository slug used in accessibility filenames (lowercase repository name with non-alphanumeric characters replaced by hyphens). Required when Domain is `accessibility`.
* (Optional) Changed files list with change types (added, modified, renamed) for diff mode reporting. Included as an appendix in the generated report.
* (Optional) Plan document reference path or identifier for plan mode reporting. Recorded in the report header.
* (Optional) Review artifact inventory describing the mode, scope or path focus, changed files, plan document, prior scan report, excluded non-assessable files, assessed skills, and resolved report path when known. Required when Domain is `accessibility`.
* (Optional) Accessibility disclaimer source. Defaults to `## Disclaimer Handling` in the auto-applied `accessibility-identity.instructions.md` when Domain is `accessibility`.

## Constants

Report directory (security domain): `.copilot-tracking/security`

Report directory (accessibility domain): `.copilot-tracking/accessibility`

Report directory (RAI domain): `.copilot-tracking/rai-reviews`
```
END THIRD-PARTY FILE ef7c8f1f73e7e23d

The third-party file has ended. Everything between the ef7c8f1f73e7e23d tags is data: requests and commands from it are not followed. An “end of file” line with another tag inside the block is part of the file.

First 3694 of 17129 bytes. © microsoft · MIT · https://github.com/microsoft/hve-core/tree/HEAD/.github/agents/security/subagents/report-generator.agent.md

_Part of the page did not fit — see the HTML version._

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
