ToolYour

Private beta, read-only: get the durable receipt for one of your saved-playbook runs.

Advanced: JSON schema for one operationId before invoke_tool. Skip this for ship/SEO/security jobs — use run_playbook or solve_task.

Private draft: return ToolYour's canonical read-only API and MCP reference, including the separate ToolYour usage boundary. It is metadata only: it does not create access, invoke an API, execute a provider operation, enable billing, or make a network request to a publisher.

Advanced: run one API-backed tool by operationId. Prefer run_playbook or solve_task for jobs. After a jobReport, apply loop.remainingFixes then verify_task — do not re-invoke the same tool.

Human-only scoped approval for one declared actionId. Requires CONTROL_PLANE_APPROVE_TOKEN. CRITICAL needs breakGlass. No approve-all.

Completion-loop: cancel an open frozen job. Does not cancel skill-loop runs (use the run lifecycle for those).

Declare a HIGH or CRITICAL host action for this job. ToolYour cannot see undeclared shell commands. Not an approve-all.

Completion-loop: start a frozen task (task-1 … task-5, or optional host-playwright). Do not use for SEO, URLs, or ship-gate — those use plan_task. Agents that already have a jobId must call job_status instead.

Completion-loop status for an existing jobId (state, frozen checks, next_action). Do not use for catalog/SEO goals — those use plan_task.

List tool category families (Convertors, Documents, SEO Tools, etc.). Use before discover_tools to narrow search — e.g. discover_tools('pdf', category: 'Documents'). Free — not billed.

Free: browse opt-in community feature patterns (anonymized). Filter by domain. Does not include private project labels.

Deprecated compatibility tool: list existing private Feature Memory records for export, review, or deletion.

Private draft: list this account's provider access requests and current access state. An approved request can later be revoked, expire, or be temporarily unavailable; this tool exposes metadata only.

Private draft: list API projects this account owns, has an active discovery grant to use, or may discover through a free metadata policy. Use one access mode with nextCursor for stable pagination. Metadata only; does not invoke provider APIs.

Private beta, read-only: list your saved ToolYour playbooks. Other users' playbooks are never returned.

List skill playbooks. Prefer run_playbook(skillId) immediately. Primary jobs: ship-gate, seo-site-audit, web-security-audit. After a run, verify_task — do not invoke_tool.

Load full skill playbook markdown. Prefer run_playbook to execute in one step.

Free skill-loop planner for supported SEO, security, ship-gate, and catalog jobs. Returns a ranked recommendation and loop decision; it does not execute the work.

Private draft: create an encrypted, idempotent invocation intent for an owned approved operation or a granted read-only operation. Granted preparation requires independent server gates and an active consumer pilot session. This does not call the provider.

Free: opt-in publish a captured feature to the anonymized community library (project/repo redacted). Others see requirements + matrix only.

Deprecated compatibility tool: recall existing Feature Memory records owned by this account. It is no longer part of the default planning path.

Beta: request human merge/release approval for a WorkRecord's immutable PR evidence. This never approves, merges, or changes GitHub.

Private draft: request access to selected approved provider operations. The provider must explicitly approve the request before any grant exists. The reason is metadata only; never include credentials or request/response content.

Run a skill playbook (ship-gate, production-readiness-gate, SEO audit, …). Read verification.evidence + loop.line first. After host fixes, verify_task with this result as baseline (or profileId). Prefer over load_skill or invoke_tool.

Private beta: run one active saved ToolYour playbook with merged input, save its receipt, and return the execution result. It reuses only registered ToolYour skills or workflows.

Run a named multi-step job by workflowId. Prefer run_playbook (skill) or solve_task (goal). Returns loop.remainingFixes; then verify_task.

Private draft: search selected and approved operation metadata this account owns, may discover through an active grant or free metadata policy, or may request from an invite-only provider. Results label whether a capability is ToolYour-owned or publisher-owned, whether it is offered through API, MCP, or both, and its availability and plan scope. Use one access mode with nextCursor for stable pagination; filter by provider, risk, tag, or category. Category is an operation classification tag. Does not expose schemas, credentials, or provider responses and does not invoke APIs.

Run a job from a plain-language goal. Workflow runs attach verification.evidence + profileId (same as run_playbook). Read loop.line — if initiate is false, stop; if true, apply rank-1 then verify_task.

Free: remove a feature from the community library while keeping it in your private memory.

Private beta: update saved-playbook details or create a new immutable version by supplying definition. Archiving prevents future runs.

Close the loop only when the prior result has loop.initiate true. Requires baseline jobReport OR profileId with a prior run_playbook snapshot. Read verification.evidence + loop.line; apply rank-1 loop.nextActions. Stops when loop.stop is set.