agentleFS
HybridAIOneHybridAIOne/hybridclaw/skills/miro/SKILL.md

miro

Skill159 starsChanged 6 months agoMITView on GitHub

Discover Miro boards, read board items for planning and summaries, prepare guarded board writes, and run Enterprise board export workflows through SecretRef-backed API requests.

  • Reads credentials

We split this skill into the rules it gives an agent

15 rules. About 0 are common, 6 are distinctive, and 9 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 9

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill for Miro board work through the REST API: board discovery, metadata lookup, board item reads for summaries or planning, …

    line 99
  2. Distinctive:

    read board items by type for audit-safe context capture and summaries

    line 107
  3. Distinctive:

    create or update sticky notes, text items, shapes, connectors, and frames only after preview and explicit operator grant

    line 109
  4. Distinctive:

    create and inspect Enterprise board export jobs, including result/task lookup and export-link creation where the API supports it

    line 111
  5. Very specific:

    Use MIRO_ACCESS_TOKEN for normal board APIs.

    line 138
  6. Very specific:

    Use MIRO_DISCOVERY_ACCESS_TOKEN only for Enterprise board export APIs that require boards:export.

    line 144
  7. Very specific:

    Do not try to verify either secret with bash, environment inspection, or by asking the model whether a secret exists.

    line 155
  8. Very specific:

    Start with list-boards unless the board id is already known.

    line 165
  9. Very specific:

    Read board items with list-items, filtered by --type when the task is about a specific surface such as sticky notes, text, shapes, …

    line 168
  10. Very specific:

    After export-results returns an exportLink, use capture-export to save the ZIP/PDF/HTML/SVG export under .generated-miro and return the …

    line 181
  11. Distinctive:

    Do not delete items, delete boards, or change board permissions through this skill.

    line 184
  12. Distinctive:

    Run the colocated helper with Node:

    line 189
  13. Very specific:

    After the operator approves the URL and stores the callback code with hybridclaw secret set MIRO_OAUTH_CODE "<code>", exchange it without …

    line 211
  14. Very specific:

    Use structured flags for common payload fields: --content, --title, --shape, --x, --y, --width, --height, --parent-id, --style-json, …

    line 277
  15. Very specific:

    Use --payload-json only when the operator has supplied a known-good Miro REST payload shape that the structured flags cannot represent.

    line 284

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Miro

Use this skill for Miro board work through the REST API: board discovery, metadata lookup, board item reads for summaries or planning, guarded creation and updates for common board items, and Enterprise board export jobs when the operator has Discovery access.

Scope

  • list accessible boards and fetch board metadata
  • read board items by type for audit-safe context capture and summaries
  • fetch specific sticky note, text, shape, connector, and frame items
  • create or update sticky notes, text items, shapes, connectors, and frames only after preview and explicit operator grant
  • create and inspect Enterprise board export jobs, including result/task lookup and export-link creation where the API supports it
  • download completed Enterprise export links into local workspace artifacts
  • classify missing credential, upstream auth/scope, rate-limit, and API errors
  • refuse deletes and permission/share changes in this v1 skill

Credential Rules

Store Miro tokens in HybridClaw runtime secrets. Never paste a raw Miro token into the prompt, a helper argument, a shell environment dump, or a log.

Recommended setup order:

  1. Browser admin: open the active HybridClaw admin URL ending in /admin/secrets and set the needed MIRO_* secrets.
  2. Browser /chat or TUI fallback: /secret set MIRO_ACCESS_TOKEN "<oauth-or-access-token>" plus any needed Discovery or OAuth client secrets.
  3. Local console fallback:
hybridclaw secret set MIRO_ACCESS_TOKEN "<oauth-or-access-token>"
hybridclaw secret set MIRO_DISCOVERY_ACCESS_TOKEN "<enterprise-discovery-token>"
hybridclaw secret set MIRO_CLIENT_ID "<miro-client-id>"
hybridclaw secret set MIRO_CLIENT_SECRET "<miro-client-secret>"

Use MIRO_ACCESS_TOKEN for normal board APIs. Required Miro scopes depend on the operation:

  • board discovery and item reads: boards:read
  • sticky notes, text, shapes, connectors, and frames: boards:write

Use MIRO_DISCOVERY_ACCESS_TOKEN only for Enterprise board export APIs that require boards:export. Enterprise export requires a Miro Enterprise plan, Company Admin role, and enabled eDiscovery.

For Miro OAuth app authorization, use the helper to build the authorize URL, store the callback code as MIRO_OAUTH_CODE, and then exchange the code through the gateway. The exchange request uses <secret:MIRO_CLIENT_ID>, <secret:MIRO_CLIENT_SECRET>, and <secret:MIRO_OAUTH_CODE> placeholders and captures access_token into MIRO_ACCESS_TOKEN and refresh_token into MIRO_REFRESH_TOKEN with captureResponseFields.

Do not try to verify either secret with bash, environment inspection, or by asking the model whether a secret exists. The model cannot inspect the gateway secret store. If the operator says the secret was set, build the helper request and pass the httpRequest object to the built-in http_request tool. Only say a secret is missing if the gateway or helper error explicitly says MIRO_ACCESS_TOKEN or MIRO_DISCOVERY_ACCESS_TOKEN is missing, unavailable, not set, or unresolved.

Default Workflow

  1. Start with list-boards unless the board id is already known.
  2. Fetch board metadata with get-board before summarizing ownership, sharing, or last-modified context.
  3. Read board items with list-items, filtered by --type when the task is about a specific surface such as sticky notes, text, shapes, connectors, or frames.
  4. For large boards, page with the returned cursor and keep summaries bounded to relevant item types and regions named by the operator.
  5. For any board write, run approval-plan or --request first, show the preview, then wait for explicit approval.
  6. Pass --operator-grant approve-miro-board-write only after approval for the exact board id, operation, and payload.
  7. For Enterprise exports, use approval-plan export-create before creating an export job and approval-plan export-link before minting a download link. Pass --operator-grant approve-miro-export only after approval for the exact org id, board id or task id, and export format.
  8. After export-results returns an exportLink, use capture-export to save the ZIP/PDF/HTML/SVG export under .generated-miro and return the helper's artifacts[] output.
  9. Do not delete items, delete boards, or change board permissions through this skill.

Command Contract

Run the colocated helper with Node:

node skills/miro/miro.cjs --help

Plan a request without contacting Miro:

node skills/miro/miro.cjs --format json plan "Summarize sticky notes on this Miro board"

Build the OAuth authorize URL:

node skills/miro/miro.cjs --format json oauth authorize-url \
  --client-id "<miro-client-id>" \
  --redirect-uri "http://127.0.0.1:1455/oauth2/callback" \
  --scope boards:read \
  --scope boards:write

After the operator approves the URL and stores the callback code with hybridclaw secret set MIRO_OAUTH_CODE "<code>", exchange it without exposing the code or resulting tokens:

node skills/miro/miro.cjs --format json http-request oauth-exchange-code \
  --redirect-uri "http://127.0.0.1:1455/oauth2/callback"

For expiring-token apps, refresh tokens through the same secret-capture path:

node skills/miro/miro.cjs --format json http-request oauth-refresh-token

Build read requests:

node skills/miro/miro.cjs --format json http-request list-boards \
  --query roadmap \
  --limit 20

node skills/miro/miro.cjs --format json http-request get-board \
  --board-id uXjVOD50NUI=

node skills/miro/miro.cjs --format json http-request list-items \
  --board-id uXjVOD50NUI= \
  --type sticky_note \
  --limit 50

node skills/miro/miro.cjs --format json http-request get-item \
  --board-id uXjVOD50NUI= \
  --type sticky_note \
  --item-id "3458764511234567890"

The helper prints { "command": "http-request", "httpRequest": { ... } }. Pass only the httpRequest value to the built-in http_request tool for live API calls. The helper sets bearerSecretName so the gateway injects the token server-side.

Preview and approve board writes:

node skills/miro/miro.cjs --format json approval-plan create-sticky-note \
  --board-id uXjVOD50NUI= \
  --content "Decision: proceed with option B" \
  --x 0 \
  --y 0

node skills/miro/miro.cjs --format json http-request create-sticky-note \
  --board-id uXjVOD50NUI= \
  --content "Decision: proceed with option B" \
  --x 0 \
  --y 0 \
  --operator-grant approve-miro-board-write

Supported write operations:

  • create-sticky-note, update-sticky-note
  • create-text, update-text
  • create-shape, update-shape
  • create-connector, update-connector
  • create-frame, update-frame

Use structured flags for common payload fields: --content, --title, --shape, --x, --y, --width, --height, --parent-id, --style-json, --position-json, --geometry-json, and --data-json. For connector creation, provide --start-item-id and --end-item-id; optional --start-snap-to, --end-snap-to, --shape, --captions-json, and --style-json map to the Miro connector request body.

Use --payload-json only when the operator has supplied a known-good Miro REST payload shape that the structured flags cannot represent. Still run approval-plan first and keep the payload minimal.

Enterprise export workflow:

node skills/miro/miro.cjs --format json approval-plan export-create \
  --org-id 3074457345821141000 \
  --board-id uXjVOD50NUI= \
  --request-id 92343229-c532-446d-b8cb-2f155bedb807 \
  --board-format PDF

node skills/miro/miro.cjs --format json http-request export-status \
  --org-id 3074457345821141000 \
  --job-id 92343229-c532-446d-b8cb-2f155bedb807

node skills/miro/miro.cjs --format json http-request export-results \
  --org-id 3074457345821141000 \
  --job-id 92343229-c532-446d-b8cb-2f155bedb807

node skills/miro/miro.cjs --format json capture-export \
  --export-url "https://..." \
  --filename "miro-board-export.zip"

Error Interpretation

  • Gateway errors saying MIRO_ACCESS_TOKEN is missing, not set, unavailable, or unresolved: ask the operator to set the runtime secret in the same HybridClaw runtime, then retry in a fresh agent runtime if the gateway was already running.
  • Gateway errors saying MIRO_DISCOVERY_ACCESS_TOKEN is missing: Enterprise export APIs cannot run until a Discovery token is stored.
  • Gateway errors saying MIRO_CLIENT_ID, MIRO_CLIENT_SECRET, MIRO_OAUTH_CODE, or MIRO_REFRESH_TOKEN is missing: the OAuth setup step has not stored the required secret in the active runtime.
  • Gateway errors saying the secret is blocked by policy: report a policy/runtime configuration problem; do not ask the operator to set the same secret again.
  • Miro 401 or 403 responses: the gateway injected a token, but Miro rejected it or the token lacks the needed board, organization, Enterprise, or OAuth scope.
  • Miro 429 responses: back off and preserve cursor/request-id values for idempotent retries.
  • Miro 5xx responses: report the upstream failure and retry only if the user asks and the operation is idempotent.

The helper can classify common failures offline:

node skills/miro/miro.cjs --format json explain-error \
  --status 403 \
  --message "insufficient scope"

References

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

Reviews can't be read right now.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.