LadderFlow

LadderFlow is a content-creation platform for founders and operators built around a Digital Brain — a structured memory of your opinions, frameworks, and…

От сообществаРаботаетБез входаГлобальныйБесплатноМожет изменять данные

Что умеет

Ladderflow Search Brain: Semantic + keyword search over the user's Digital Brain (their extracted opinions, frameworks, stories, proof points, beliefs). Each result carries a `source` field: 'user' =
Ladderflow Get Memory Item: Fetch one Brain memory in full — its content, trust tier, edit history, which posts it contributed to, and (for reference-sourced memories) the external source it came from
Ladderflow Get Brain Summary: A high-level summary of the user's Digital Brain: how many memories they have, broken down by type and trust tier, plus their most-discussed topics (top_topics, ordered b

Какие данные видит

Нужен ли аккаунт

Не нужен: сервер работает без входа

LadderFlow is a content-creation platform for founders and operators built around a Digital Brain — a structured memory of your opinions, frameworks, and stories, built primarily through short voice interviews. Through this MCP server, an AI assistant can search and add to your Brain, run the full voice-interview funnel, generate and edit LinkedIn/X drafts from your own material, and check your post performance — all read/generate/draft only; publishing always stays a human, in-app action.

Список инструментов сервера (20)

Технические названия из tools/list. Нужны только разработчикам.

ladderflow_search_brain	Semantic + keyword search over the user's Digital Brain (their extracted opinions, frameworks, stories, proof points, beliefs). Each result carries a `source` field: 'user' = the user's own material, 'reference' = external source material they saved (cite it as external, never as the user's own words); and a `trust_tier` of A, B or C (A is the most trusted; only A and B feed generation).
ladderflow_get_memory_item	Fetch one Brain memory in full — its content, trust tier, edit history, which posts it contributed to, and (for reference-sourced memories) the external source it came from.
ladderflow_get_brain_summary	A high-level summary of the user's Digital Brain: how many memories they have, broken down by type and trust tier, plus their most-discussed topics (top_topics, ordered by how many times each was discussed — use this, not list_topics, for 'most-discussed' or 'most talked about' questions). Good first call to understand what the Brain holds.
ladderflow_list_topics	List the topics the user has discussed, most recent first. For the user's MOST-DISCUSSED or most-frequent topics, use get_brain_summary instead (its top_topics list) — this tool sorts by recency, not frequency.
ladderflow_add_to_brain	Add material to the user's Digital Brain. Provide `text` (a thought to save as a memory) and/or `source_url` (any link to ingest — YouTube, LinkedIn, X, a direct PDF link, or a general web article). At least one is required.
ladderflow_get_reference_status	Check on a URL you added with add_to_brain. Pass the reference_source_id it returned. Tells you whether ingestion is still running, completed (and how many memories it produced), or failed (with a reason). Use this to confirm a URL add actually landed before searching for it.
ladderflow_list_sessions	List the user's voice interview sessions, most recent first.
ladderflow_get_session_status	Get one interview session's status/phase, how many Brain memories it produced, and (if available) its running transcript so you can summarize what's been discussed so far. Accepts either a real session id or the scheduled_interview_id returned by launch_interview — before the user has clicked that link, this reports a 'not started yet' status instead of an error.
ladderflow_list_drafts	List the user's generated posts/drafts, most recent first. Includes remix and MCP-generated drafts (which the in-app list can miss). Each row shows lifecycle `status` and `publish_status` so you can tell unpublished drafts from published posts.
ladderflow_get_draft	Fetch one draft/post in full by its id (works for remix and MCP-generated drafts too). READ-ONLY — this only returns the text, it never changes it. To change/rewrite a draft, call update_draft with the full new text (often after reading it with this tool first).
ladderflow_get_review_link	Get a secure link into the LadderFlow app to review/edit/publish a draft (publishing always happens in-app behind human review — never over MCP). With a draft_id, links to that draft's review page (session review for interview-backed drafts; /posts?id= for interview-less MCP/remix drafts); without one, links to the posts queue.
ladderflow_update_draft	Edit or rewrite a draft's content — use when the user wants to change, tighten, shorten, punch up the hook/opening, or otherwise revise a draft (you must supply the full new content). Saves the edit (which also feeds the app's edit-learning loop). Editing never publishes — publishing stays in-app. NOT for reading a draft (use get_draft) or generating new posts from scratch (use generate_posts).
ladderflow_get_performance	The user's content performance for a period (this_week / last_week / last_30d): published post count, impressions, best post, and per-post metrics/ratings. Pro feature.
ladderflow_get_friday_review	The user's weekly Friday Review: this week's published posts with their ratings/outcomes and pipeline results (DMs, calls, deals, revenue). Pro feature.
ladderflow_get_winning_patterns	The user's best-performing post structures/hooks (by average rating) plus a short brief on what's been working — useful before writing new posts. Pro feature.
ladderflow_generate_posts	Generate draft posts from the user's Digital Brain. Provide EXACTLY ONE of: `topic` (pull relevant memories), `memory_item_ids` (use specific memories), or `signal` (a ready-made signal dict). Creates unpublished drafts (never publishes) — review/publish in-app via get_review_link.
ladderflow_remix_post	Create a fresh draft by remixing an existing post — reuses the Brain memory that post was built on to write a new angle in the same format. Pro feature. (If the post has no linked memory, you'll get a list of remixable high-performers to try instead.) Source must be an original post, not a remix.
ladderflow_start_interview	Kick off research for a voice interview on a topic. Returns a research_id immediately; research runs in the background (~5–7 min). Then call get_interview_angles with that id to see candidate angles.
ladderflow_get_interview_angles	Check a research run. Returns 'still researching' until it's done, then candidate angles (each with a ready outline). Show the angles to the user; when they pick one, call launch_interview.
ladderflow_launch_interview	Launch the live voice interview for a chosen angle. Pass the research_id and the angle_id the user picked (optionally an edited_outline to override title/global_context/why_this_matters/key_questions). Returns a secure deep link that starts the interview in the user's browser — no room or token is exposed here. Only call this after get_interview_angles reports status complete and you have a real angle_id — calling earlier returns an error.