Skip to content
Back to skills

Miro Debug Bundle

ASecurity

Design and implement a repository-side Miro diagnostic-bundle workflow with bounded metadata and automatic redaction. Use when escalating a Miro failure. Trigger with \"build Miro debug bundle\".

  • 2,719 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added September 21, 2026
ai-agentsdebuggingapidatabasesecuritydocumentation

Works with

  • claude code
  • cli
  • api

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned September 21, 2026

npx -y skills add jeremylongshore/claude-code-plugins-plus --skill miro-debug-bundle --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Miro Debug Bundle?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Miro Debug Bundle
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/jeremylongshore-miro-debug-bundle-a42104dc/badge)](https://www.skillsdirectory.com/skills/jeremylongshore-miro-debug-bundle-a42104dc)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: miro-debug-bundle
description: "Design and implement a repository-side Miro diagnostic-bundle workflow with bounded metadata and automatic redaction. Use when escalating a Miro failure. Trigger with \"build Miro debug bundle\"."
argument-hint: "[incident-id] [time-window]"
allowed-tools: Read, Glob, Grep, WebFetch, Write, Edit
version: 1.9.0
author: Jeremy Longshore <jeremy@intentsolutions.io>
license: MIT
tags:
- saas
- miro
- debugging
- support
model: inherit
effort: medium
compatibility: Designed for Claude Code; live work requires an authorized Miro app and redacted evidence
---
# Miro Redacted Debug Bundle

## Overview

Create evidence useful to maintainers or Miro support while excluding credentials, authorization codes, board content, personal data, and unrestricted configuration; use the evidence produced here to make the next decision explicit and reviewable.

## Prerequisites

- Incident identifier and narrow UTC window
- Approved evidence destination and retention period
- Known redaction rules for identifiers, URLs, headers, and payloads

## Tool Discipline

Use `Read`, `Glob`, and `Grep` to inspect the repository, configuration names, adapters, tests, and evidence. Use `WebFetch` only for current official Miro documentation. Use `Write` or `Edit` after confirming the requested mode, target environment, tenant, board, and approval boundary. These declared tools do not call authenticated Miro APIs or deployment CLIs; implement client, configuration, and test changes, then return exact operator commands or an approval-gated handoff for live execution.

## Current Contract

- Safe evidence includes versions, operation names, status/code, latency, rate headers, counts, hashes, and timestamps.
- Authorization headers, cookies, tokens, client secrets, callback query strings, and raw board payloads are never safe evidence.
- Access-token context is useful only after user/team identifiers are pseudonymized.
- Official status evidence should include observation time because incidents change.

## Authentication

For REST work, use OAuth 2.0 Authorization Code with the narrowest Miro scopes. Bind each encrypted token record to its user, application, authorized team, and granted scopes. Never print access tokens, refresh tokens, client secrets, authorization codes, or board content.

## Instructions

1. Define the incident window, failing operations, recipient, purpose, and retention.
2. Collect repository version, deployment version, app mode, safe configuration names, and dependency lock facts.
3. Extract bounded request metadata and rate headers; replace identifiers with stable incident-local hashes.
4. Record token-context match as a boolean and include no credential or raw context response.
5. Scan the bundle for bearer patterns, secrets, emails, callback codes, board text, and private URLs.
6. Review the manifest and redaction report before writing the approved artifact.

## Approval Boundaries

Do not collect raw payloads, full environment/configuration dumps, database exports, or user content without data-owner and security approval. Pause when the responsible owner or exact target is uncertain.

## Output

Return bundle path, manifest, time window, redaction counts, evidence gaps, sensitivity label, recipient, and deletion date. State what was not inspected or changed so the receipt cannot overclaim coverage.

## Error Handling

| Condition | Response |
|---|---|
| Redaction scanner matches | Quarantine the bundle and regenerate from safer fields. |
| Evidence source exceeds scope | Skip it and document the gap. |
| Timestamps disagree | Normalize to UTC and preserve original offsets. |
| Destination is not approved | Do not write or upload the bundle. |

## Examples

The example is a redacted operator receipt; identifiers are hashes or bounded labels, not board content or credentials.

```text
incident=INC-204; window=18m; requests=14; ids-hashed=9; secrets-found=0; content-fields=0; retention=7d
```

## Resources

- [Skill-specific official documentation](references/official-docs.md)
- [Security guidelines](https://developers.miro.com/docs/security-guidelines)
- [Rate limiting](https://developers.miro.com/reference/rate-limiting)

Files in this skill

  • SKILL.md4.1 KB
  • references/official-docs.md986 B

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.