Turn scattered ideas into finished work — pergunta.me
Turn scattered ideas into finished work — pergunta.me
Pergunta turns overwhelming to-dos into a realistic plan with a clear next step — and adapts when your day shifts. Start for free, no credit card.
This listing identifies a third-party MCP server. MCPBundles is not affiliated with, sponsored by, or endorsed by the service or its operator unless stated on this page.
This server may ask you to sign in when your AI client connects.
Last probed Nov 3, 2026 · mcp.pergunta.me
35tools discovered
For AI agents — MCPBundles Directory API
For MCP client apps (human connect):
MCP connect URL: https://mcp.pergunta.me/mcp
Skill page: https://www.mcpbundles.com/skills/pergunta-mcp-4d2c2f4134
Explore capabilities (35)
read 11
write 14
Showing 25 of 35
Showing 25 of 35 tools from the public preview. Connect the server to inspect the full live tool list.
Get5
List3
Add Task2
Attach Task2
Other capabilities13
Capture Tasks
write
capture_tasks
Batch create tasks from conversation. Use when user mentions multiple things they need to do. Tasks are created in inbox status.
Capture Tasks — Batch create tasks from conversation. Use when user mentions multiple things they need to do. Tasks are created in inbox status.
Review Task Capture — Review a ChatGPT task-capture draft before saving. Returns rewrites, enhancements, clarifying questions, and a save decision without persisting the task.
Accept Negotiated Capture — Commit a negotiated task-capture proposal returned by review_task_capture, optionally applying per-field user overrides. Persists negotiation provenance to the audit log.
List Tasks — List user tasks with optional filters. Supports offset-based pagination via limit and offset parameters; the response includes has_more to indicate additional pages. When status is omitted and no active tasks exist, the response falls back to all statuses and includes status_hint so subsequent pages can pass it explicitly. Returns tasks ordered by manual task order — except when focused=true, which orders by focused_at DESC (most-recently-pinned first) and never falls back to status:all, even on
Search Tasks — Search tasks by free-text query against task titles, notes, and linked knowledge (comments, indexed chunks), with minimal pagination over matching results. Use this for content/keyword search; use list_tasks instead when filtering by status, priority, or parent without a text query.
Get Task — Return a single task. Default response is LEAN: title, summary, tags, status, priority, parent/breadcrumb, dates, subtask_count, and orientation fields. Pass `fields` to include one or more optional detail sections for THIS task only (notes, comments, linked_resources, dynamic_fields, knowledge). For tasks with large comment histories use comments_limit (1–100) and comments_cursor (from commentPagination.next_comment_cursor) to paginate through comments in bounded, insert-safe pages. commentPagi
Get Next Leaf Task — Return the next eligible unblocked task matching the required status. A task is eligible when it has no blocker_reason and is a leaf, or when all of its leaf descendants are done. Candidates can be restricted to descendants under parent_task_id and/or tasks matching all tag_ids, then are sorted by pinned/focused first, priority, and manual sort_order.
Get your MCP into directories
A working endpoint is step one. Directory coverage is the coordinated launch across ChatGPT, Claude, Cursor, the MCP Registry, and community indexes.
What is the Turn scattered ideas into finished work — pergunta.me MCP server?
Pergunta turns overwhelming to-dos into a realistic plan with a clear next step — and adapts when your day shifts. Start for free, no credit card.
How do I connect Turn scattered ideas into finished work — pergunta.me to my AI agent?
Use the MCP endpoint listed on this page in your MCP client configuration. One-click install pills support Claude, Cursor, VS Code, and other hosts. Copy the remote MCP URL if your client needs a manual entry.
Operate this server?
Request removal if you do not want Turn scattered ideas into finished work — pergunta.me listed here, or claim the listing if you want to manage it on MCPBundles.
Operate Turn scattered ideas into finished work — pergunta.me?
This server appears in the MCPBundles directory. Verify you operate it to take over the listing — name, description, logo, contact email, and skill content. We email a 6-digit code to a maintainer address your server publishes in /.well-known/security.txt or /.well-known/mcpbundles.json. Free, takes about a minute.
MCPBundles publishes this directory for MCP discovery. Except where we host or operate an offering, third-party services run under their own terms. Product and company names on this page are used in a descriptive, identifying way (including under nominative fair use where applicable); they remain the property of their owners. Nothing here grants you rights in those marks, and nothing here is an offer to sell a third party's services. Terms · Request removal
Update Task — Update an existing task. Provide task_id and any fields to update. Only the specified fields will be changed. Tags can be edited natively: add_tags/remove_tags add or remove tags (combinable), or set_tags replaces the whole tag set (mutually exclusive with add_tags/remove_tags). Each tag reference is an owned tag UUID or a tag name; unknown names are auto-created, unknown UUIDs and ambiguous names are rejected without blocking the rest of the update. Protected execution-context fields (brief, de
Complete Task — Mark a task as complete (status=done). Sets completed_at timestamp automatically.
Add Task Comment — Add a comment to a task. Use this when the user wants to add a comment or note to a specific task.
Add Task Link — Add a link (bookmark) to a task, visible in the Pergunta web app's Links panel. This is NOT an agent-readable URL store: get_task/list_tasks/search_tasks always return source_url as null for links (URL must never reach the LLM via MCP), so no future MCP call (including by the same caller) can retrieve the URL back. If you need the URL to remain readable by an agent later, use add_task_comment instead.
Attach Task Photo — Attach a photo to an existing task, reusing the exact same encrypted/EXIF-scrubbed/quota-enforced pipeline the Pergunta REST photo-attach endpoint uses. dataUrl is a base64 data URL (not multipart/binary), max ~13.3MB encoded (10MB decoded) — this is primarily useful for programmatic/agentic MCP clients, not something a chat LLM types by hand given the payload size. Unknown extra arguments are silently ignored (not rejected) — an inherited REST-schema characteristic, not a bug.
Attach Task File — Attach a generic file (PDF/DOC/XLS/PPT/TXT/CSV/MD) to an existing task, reusing the exact same encrypted/quota-enforced pipeline the Pergunta REST file-attach endpoint uses. dataUrl is a base64 data URL (not multipart/binary), max ~13.3MB encoded (10MB decoded) — this is primarily useful for programmatic/agentic MCP clients, not something a chat LLM types by hand given the payload size. mimeType and sizeBytes are REQUIRED (unlike attach_task_photo, where they are optional/inferred). Unknown extr
List Task Artifacts — List the artifacts attached to a task: file/photo attachments and link bookmarks. Read-only, no side effects. Attachment items include a preview_url — a bare, same-origin, auth-gated relative path (not a signed URL); it requires the same caller credential to fetch, identical to the REST attachments endpoint. Link items are ALWAYS redacted: source_url and raw_content are null (URL must never reach the LLM via MCP), same policy as get_task/list_tasks/search_tasks/add_task_link. Each section is cap
Get Task Risk Routing — Return the latest risk-routing assessment for an owned task: route, five factor scores and reasons, named overrides, and the deterministic explanation. Read-only. assessment is null when none is stored. Does not propose or persist a route.
Get Execution Context — Read this context before acting. Stop if readiness is blocked or confirmation is required. Return the canonical execution snapshot for one initiative. Primary selector is a canonical UUID (root_task_id). Default view is concise; pass view=full for decisions and brief fields. Always includes version, readiness, blockers, risk, and verification. Read-only — never mutates.
Get Next Executable Action — Read this context before acting. Stop if readiness is blocked or confirmation is required. Return the current executable action plus readiness, blockers, risk, verification, and context version. UUID primary selector. Read-only — never mutates. Stop if blocked.
List Open Decisions — Read this context before acting. Stop if readiness is blocked or confirmation is required. List open decisions on an initiative together with readiness, blockers, risk, verification, and context version. UUID primary selector. Read-only — never mutates. Do not act while an open decision blocks confirmation.
Find Initiatives — Read this context before acting. Stop if readiness is blocked or confirmation is required. Ambiguity-safe initiative search. Returns candidates only; selected is always null, even on a unique match. Never auto-selects a write or read target. Copy a canonical UUID into get_execution_context. Read-only.
Submit Support Report — Submit product feedback or a bug report. Include only what the user deliberately wants to send; never copy task text or page content.
Claim Task — Atomically acquire a lease on a task for a worker. Succeeds (won=true) only if the task is unclaimed, its lease has expired, or the caller proves continuity by passing BOTH expected_owner (matching the current claim_owner) AND expected_claim_token (matching the current fencing token). expected_owner alone does NOT supersede an unexpired lease — that would be lease theft. Contention (a validly-held task) returns won=false. On success returns a fencing claim_token and the lease expiry (lock_until)
Acquire Next Leaf Task — Find eligible unblocked tasks with the same ordering as get_next_leaf_task, then atomically acquire the first claimable task lease. Contended tasks are skipped up to 10 ordered candidates per call.
Renew Claim — Heartbeat: extend the lease on a task the caller already holds. Succeeds (renewed=true) only if claim_owner AND claim_token still match. renewed=false means the claim was superseded by another worker — the caller must stop work immediately. ttl_seconds is clamped server-side to [1, 7200].
Release Claim — Release a lease on a task on clean finish or abort. Succeeds (released=true) only if claim_owner AND claim_token still match. released=false means the lease is no longer the caller's. The fencing claim_token is preserved (never reset) so it can never be reused.
Verify Claim — Cheap, read-only fencing check before a destructive operation (merge, status->done). valid=true only if the caller still owns the lease, the claim_token matches, and the lease has not expired. Use immediately before any irreversible action to prevent a slow worker from acting after its lease was stolen or lapsed.
Showing the first 25 tools — full list and schemas: Directory API.
How many tools does Turn scattered ideas into finished work — pergunta.me provide?
MCPBundles probed 35 tools on the live server. The tool list on this page reflects what was discovered at the last refresh — connect your client to see the full set available to your session.
What authentication does Turn scattered ideas into finished work — pergunta.me require?
Turn scattered ideas into finished work — pergunta.me may require signing in to the provider before tools can run. Connect through MCPBundles or your MCP client and complete any provider login when prompted.
Don't want this listed?
Request removal from the MCPBundles directory. We stop publishing this listing on mcpbundles.com.
MCPBundles is an independent platform built on the open Model Context Protocol standard. MCPBundles is not affiliated with, sponsored by, or endorsed by Anthropic PBC.