Build on Outcome from your own tools.
Outcome is reachable three ways: a connector for AI assistants, an API for embedding, and an app for Atlassian. This page lists what each one serves.
One address, four permissions. Live
Any assistant that supports the Model Context Protocol connects at the address below and signs in as a person. The person chooses which permissions the connection carries, and every call is logged for the company's administrators.
https://outcomefactory.ai/mcpclaude mcp add --transport http outcome https://outcomefactory.ai/mcpread_spine | The goal tree, with each goal's measure and the share no work covers. |
|---|---|
summarise_goal | Where one goal stands, for a leadership review. |
read_brief | Why a goal exists and what building under it must respect. |
read_work_context | What a pull request is for and which number shows it working. |
read_readings | The dated figures behind a measure. |
read_work | Work in flight, what goal it serves, and what is blocked. |
read_my_work | The asker's own goals, decisions waiting and AI use. |
read_signals | Customer themes, sized, with what was decided before. |
read_behaviour | What users do in the product, and where they drop out. |
read_judgments | Decisions people confirmed or corrected, to learn from. |
read_people | Who owns what, and who should decide. |
read_ai_spend | AI spend by tool, model and team. |
read_portfolio | Initiatives, with what is in or out of the plan. |
read_capacity | Each team's capacity against its demand, by period. |
read_capacity_history | Where a team's capacity went last quarter, and what it says about the next. |
read_report | The figures for a weekly update, a review or a board paper. |
read_report_snapshot | A saved set of report figures, to quote exactly. |
check_report | Checks a drafted report's figures against the record. |
show_card | A chart or card, with every figure filled by Outcome. |
search_docs | Finds the help page for a question. |
read_doc | Reads one help page. |
list_import_sources | The sources a file of work can be loaded into. |
read_import_report | What an import loaded, and what it refused. |
start_trace_run | Opens an analysis that findings are recorded against. |
|---|---|
record_finding | Records one conclusion, with its evidence. |
finish_trace_run | Closes the analysis with a short summary. |
propose_outcome | Drafts a goal for a person to accept. |
propose_restatement | Suggests a change to a goal's wording or measure. |
propose_attribution | Suggests which goal a piece of work serves. |
propose_breakdown | Drafts the work under a goal. |
propose_scenario | Tests a change to the plan against capacity. |
draft_standing_card | Drafts a card to post on a schedule, for a person to sign. |
note_learning | Tells Outcome where the assistant could not answer well. |
record_reading | Files a dated figure a person gave, on a measure kept by hand. |
|---|
import_work | Loads work from an export, with a preview first. |
|---|
The embed API
Read a company's outcomes into your own product. Preview
A read-only API, with keys scoped to one organisation. Its specification is public.
https://outcomefactory.ai/api/openapi.jsonThe Atlassian app
Goal cards in Jira and Confluence. Preview
A Forge app places a team's goals beside its work in Jira and Confluence.
A plain-text page tells an assistant how to prepare a company's first draft and what to do once connected.
https://outcomefactory.ai/goal/start/setup