ROADMAP — Impact Model Specification — Go-first model, JSON Schema, measurement, and leadership reporting¶
Initiative: INIT-THEORYOFCHANGESPEC-001
Repository: github.com/ProductBuildersHQ/impact-model
Status: In progress — 20 of 23 phases completed
RMI IDs are stable and permanent. Commits implementing an item carry its own ID as the trailer,
Refs: <RMI-ID>. Items from before the rename to impact-model areRMI-THEORYOFCHANGESPEC-001to-076; later items areRMI-IMPACTMODEL-<NNN>. Phase status is derived from member RMIs — a phase is complete only when all its RMIs are complete. RMI0NNcorresponds to PLAN stepP-NN; phases order the work, so numbers are not sequential across phases. Phases 1–5 are the thin slice (adopter milestones M0 and M1). Requirements in eachSatisfiesline are defined in PRD.md and TRD.md.
Phase 1 — Core Model and Schema¶
Theme: Scaffold the module, map the Theory of Change vocabulary to Go types with a small minimal-valid document, and publish a lint-clean, embedded JSON Schema Status: Completed — 6 of 6 items completed
- [x]
RMI-THEORYOFCHANGESPEC-001Repository scaffold: go.mod, CI workflows (go-ci, go-lint, go-sast), lint config, README, CLAUDE.md, LICENSE - Type: enabler · Status: completed
- Satisfies: — (enabler)
- Acceptance:
go build ./...andgolangci-lint runpass; CI workflows green - [x]
RMI-THEORYOFCHANGESPEC-002Concept-to-type mapping and glossary from the research brief - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-001 - Satisfies: PRD G-1
- Acceptance: every core concept of research §8.4 and §9.8 maps to a planned Go type or a documented deferral
- [x]
RMI-THEORYOFCHANGESPEC-003pkg/refreference grammar (model:<doc>#<element>) with slug validation - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-001 - Satisfies: PRD-FR-604, TRD-015
- Acceptance: parse/format round-trip; invalid refs rejected
- [x]
RMI-THEORYOFCHANGESPEC-004pkg/impactcore document model with a minimal required set (goal, outcome, link, indicator), reusing prism-core - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-002,RMI-THEORYOFCHANGESPEC-003 - Satisfies: PRD-FR-101…110, TRD-001, TRD-016
- Acceptance: round-trip marshal tests; enum
Valid/Allhelpers tested; minimal fixture round-trips - [x]
RMI-THEORYOFCHANGESPEC-005ObservationSet, Status, StatusDiff, and ActorRegistry document types - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-004 - Satisfies: PRD-FR-111, PRD-FR-406, TRD-002
- Acceptance: round-trip tests for each document kind
- [x]
RMI-THEORYOFCHANGESPEC-006Schema pipeline:cmd/generate-schema, embeddedschema/with Get/Names/Validate, schemakit lint, CI drift check - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-004,RMI-THEORYOFCHANGESPEC-005 - Satisfies: PRD-FR-201, PRD-FR-202, TRD-003, TRD-004
- Acceptance:
schemakit lint --property-case camelCasereports 0 violations; regeneration leaves no diff; every schema loads from embed
Phase 2 — Load, Validate, and Edit¶
Theme: Trusted documents, a queryable causal graph, and one validated patch engine through which every edit flows Status: Completed — 6 of 6 items completed
- [x]
RMI-THEORYOFCHANGESPEC-007pkg/impactioJSON/YAML loading with limits, observation-set merge, and canonical formatting - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-006 - Satisfies: PRD-FR-203, PRD-FR-704, TRD-005, TRD-031
- Acceptance: one fixture in YAML and JSON yields equal documents;
Formatidempotent - [x]
RMI-THEORYOFCHANGESPEC-008pkg/validateschema, ID uniqueness and format, reference resolution with JSON-pointer issues - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-007 - Satisfies: PRD-FR-301, TRD-006
- Acceptance: table tests per error class
- [x]
RMI-THEORYOFCHANGESPEC-009pkg/graphbuild and queries (upstream, downstream, paths to impact, cycles, levels, orphans, dead ends, ceiling) - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-008 - Satisfies: PRD-FR-302, TRD-007
- Acceptance: unit and property tests; 1,000-node benchmark recorded
- [x]
RMI-THEORYOFCHANGESPEC-046pkg/editpatch engine: RFC 6902 with@idaddressing, atomic apply, error-delta rejection, dry run - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-008 - Satisfies: PRD-FR-701, PRD-FR-702, PRD G-14, TRD-029
- Acceptance: error-introducing patches rejected with input untouched; unknown
@idfails the whole patch - [x]
RMI-THEORYOFCHANGESPEC-011pkg/serviceandcmd/impact-model(Cobra) with validate, graph, schema, init, fmt, and patch commands - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-009,RMI-THEORYOFCHANGESPEC-046 - Satisfies: PRD US-1, PRD US-16, TRD §6.1
- Acceptance: CLI tests cover exit codes and output formats;
fmt --checkexits 1 on unformatted input - [x]
RMI-THEORYOFCHANGESPEC-047Element commands (add/set/rm --cascade) compiled to patches for core element kinds - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-046,RMI-THEORYOFCHANGESPEC-011 - Satisfies: PRD-FR-703, PRD US-16, TRD-030
- Acceptance: each command equals its patch equivalent;
rmwithout--cascaderejects dangling references
Phase 3 — Pathways, Metric Library, Provenance, and Reference Examples¶
Theme: Program pathways, shared metric definitions, a glossary, and provenance on every element, proven on fictional reference examples Status: In progress — 4 of 5 items completed
- [x]
RMI-THEORYOFCHANGESPEC-032Pathways (program areas) and parent / child ToC references, with element commands - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-008,RMI-THEORYOFCHANGESPEC-047 - Satisfies: PRD-FR-112, PRD US-9, TRD-017
- Acceptance: pathway round-trip; child ToC outcomes resolve against the parent
- [x]
RMI-THEORYOFCHANGESPEC-033Metric library, glossary, and indicator extensions (metricId, pathwayIds, claim, caveats, rollupFrom, theme_count), with element commands - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-032 - Satisfies: PRD-FR-113, PRD-FR-114, PRD-FR-117, PRD US-10, TRD-017
- Acceptance: roll-up declarations validated against metric aggregation; claim defaults to contribution when omitted
- [x]
RMI-THEORYOFCHANGESPEC-052Provenance on every element (stated / inferred / hypothesis / confirmed / rejected, cited evidence), indicator dataAvailability, and Consultation records; rejected elements excluded from analysis - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-033 - Satisfies: PRD-FR-124, PRD-FR-125, PRD-FR-123, TRD-036
- Acceptance: round-trip tests; rejected elements absent from graph, measurement, and default reports
- [x]
RMI-THEORYOFCHANGESPEC-010Reference examples (fictional) with provenance: intermediary organization (three pathways, two impact layers, metric library, glossary) and market adoption (open agent-authorization standard, AAER categories) - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-033,RMI-THEORYOFCHANGESPEC-052 - Satisfies: PRD G-1, PRD G-9, PRD US-1
- Acceptance: both examples validate; exercise groups, feedback links, qualitative indicators, and every provenance basis
Phase 4 — Workshop Pack and Source Drafting¶
Theme: Draft a theory from public sources and render the pathway diagram, logic models, metric definitions, glossary, CSV exports, and a Marp workshop deck from the document alone — adopter milestone M0 Status: Completed — 6 of 6 items completed
- [x]
RMI-THEORYOFCHANGESPEC-022Versionedreport.Viewmodel built from document and graph, status optional, with basis badges - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-009,RMI-THEORYOFCHANGESPEC-010 - Satisfies: PRD-FR-501, PRD-FR-502, PRD-FR-505, TRD-010
- Acceptance: view tests with and without a status snapshot; inferred elements carry badges
- [x]
RMI-THEORYOFCHANGESPEC-023Mermaid pathway renderer with level subgraphs, optional health styling, feedback links, ceiling, and pathway filter - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-022 - Satisfies: PRD-FR-501, PRD G-12
- Acceptance: golden Mermaid output
- [x]
RMI-THEORYOFCHANGESPEC-037Logic-model (per pathway), metric-definitions, and glossary reports - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-022 - Satisfies: PRD-FR-505, PRD US-12, TRD-022
- Acceptance: golden files over the intermediary example
- [x]
RMI-THEORYOFCHANGESPEC-049pkg/slidesMarp engine and workshop deck (impact-model deck workshop) - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-023,RMI-THEORYOFCHANGESPEC-037 - Satisfies: PRD-FR-801, PRD G-15, PRD US-17, TRD-033
- Acceptance: golden workshop deck; deck numbers and labels equal the reports'
- [x]
RMI-THEORYOFCHANGESPEC-039impact-model reportdefinition kinds,definitionspack, andimpact-model export csv(indicators, metrics, glossary) - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-023,RMI-THEORYOFCHANGESPEC-037,RMI-THEORYOFCHANGESPEC-011 - Satisfies: PRD-FR-204, PRD-FR-505, PRD G-12, TRD-021
- Acceptance: CLI tests; CSV golden files (RFC 4180)
- [x]
RMI-THEORYOFCHANGESPEC-053impact-model evidence addand the impact-draft-from-sources agent skill - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-047,RMI-THEORYOFCHANGESPEC-052,RMI-THEORYOFCHANGESPEC-039 - Satisfies: PRD-FR-606, PRD US-19, PRD G-17, TRD-037
- Acceptance: smoke test over saved fixture pages yields a draft where every element has a basis and every
statedelement cites a source
Phase 5 — Review and Agreement¶
Theme: Assessment report, Pandoc-ready sharing exports with stamps, review packet and sheet, review import as proposals, and approval of the agreed baseline — adopter milestone M1 Status: Completed — 5 of 5 items completed
- [x]
RMI-THEORYOFCHANGESPEC-054Assessment report: provenance mix, gaps, data availability, assumptions needing testing, generated questions, short narrative - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-022,RMI-THEORYOFCHANGESPEC-052 - Satisfies: PRD-FR-512, PRD US-19, TRD-038
- Acceptance: golden file over the intermediary example
- [x]
RMI-THEORYOFCHANGESPEC-056Pandoc Markdown profile with PNG diagrams, export stamps on every artifact, andimpact-model export docx|pptx - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-039,RMI-THEORYOFCHANGESPEC-049,RMI-THEORYOFCHANGESPEC-023 - Satisfies: PRD-FR-901, PRD-FR-902, TRD-039
- Acceptance: golden Markdown; stamp present in every format; pandoc export test skipped when absent
- [x]
RMI-THEORYOFCHANGESPEC-057Review packet (ID-tagged response tables) and review sheet (CSV) exports - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-054,RMI-THEORYOFCHANGESPEC-056 - Satisfies: PRD-FR-903, PRD US-20, TRD-039
- Acceptance: golden packet and sheet; every element row carries its stable ID
- [x]
RMI-THEORYOFCHANGESPEC-058ReviewResponse schema andimpact-model import review(JSON and --from-csv) producing proposed patches and a consultation record - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-057,RMI-THEORYOFCHANGESPEC-046 - Satisfies: PRD-FR-904, PRD-FR-905, PRD US-20, TRD-040
- Acceptance: seeded rows from every origin map to ops or flagged notes; stale hashes flagged; document never modified
- [x]
RMI-THEORYOFCHANGESPEC-055impact-model approvewith approvers, revision, version label, and refusal checks - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-058,RMI-THEORYOFCHANGESPEC-052 - Satisfies: PRD-FR-707, PRD US-21, TRD-041
- Acceptance: approval sets status and revision through the patch engine; refuses on errors or open hypotheses without --force
Phase 6 — Measurement Framework Design¶
Theme: Metric selection and collection burden, audiences and stories, UNDG change analysis, and the full intermediary example — adopter milestone M1.5 (collection plan) Status: Completed — 5 of 5 items completed
- [x]
RMI-THEORYOFCHANGESPEC-034Metric selection criteria and scores, collection cost, measurement budget, andpkg/burdentotals - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-033 - Satisfies: PRD-FR-115, PRD G-10, PRD US-11, TRD-020
- Acceptance: burden arithmetic table tests (minutes per week, questions per respondent per year, per pathway and mode)
- [x]
RMI-THEORYOFCHANGESPEC-035Audiences, theme codebook, and Story / LearningReview observation types - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-033 - Satisfies: PRD-FR-116, PRD-FR-117, PRD-FR-118, TRD-017, TRD-018
- Acceptance: unknown theme IDs and missing consent rejected by validation
- [x]
RMI-THEORYOFCHANGESPEC-044Change-analysis model: problem-tree causes, outcome responsibility and delivering actors, prioritization criteria for outcomes and interventions, target groups, assumption and risk enrichment - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-032,RMI-THEORYOFCHANGESPEC-034 - Satisfies: PRD-FR-105, PRD-FR-108, PRD-FR-119…122, TRD-025
- Acceptance: round-trip tests; unknown cause / target-group / actor references rejected by validation
- [x]
RMI-THEORYOFCHANGESPEC-036Intermediary example extended with framework-design and change-analysis data; schemas regenerated lint-clean - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-034,RMI-THEORYOFCHANGESPEC-035,RMI-THEORYOFCHANGESPEC-044 - Satisfies: PRD G-9 success metric
- Acceptance: example with 8–12 selected metrics, budget, audiences, codebook, sample stories, problem tree, partner-delivered outcomes, and consultations validates;
schemakit lint0 violations - [x]
RMI-THEORYOFCHANGESPEC-038Metric-selection and collection-plan reports and CSV exports - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-022,RMI-THEORYOFCHANGESPEC-034,RMI-THEORYOFCHANGESPEC-039 - Satisfies: PRD-FR-505, PRD US-11, PRD US-12, TRD-022
- Acceptance: golden files; collection-plan totals equal
pkg/burdenresults
Phase 7 — Quality: Lint, Rubric, and Judging¶
Theme: Deterministic lint with profiles, the LLM-as-a-Judge rubric with a development profile, and reproducible file-based judging with multi-judge agreement Status: Completed — 6 of 6 items completed
- [x]
RMI-THEORYOFCHANGESPEC-012Lint engine: rule registry, draft / standard / full profiles, options, structured-evaluation SummaryReport output - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-009 - Satisfies: PRD-FR-303, PRD-FR-304, PRD-FR-804, TRD-008, TRD-035
- Acceptance: registry and report-shape tests; minimal fixture is GO under
draft - [x]
RMI-THEORYOFCHANGESPEC-013Structure and logic rules (TOC-STR-, TOC-LOG-) with seeded-defect fixtures - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-012 - Satisfies: PRD-FR-303, TRD §6.4
- Acceptance: each rule triggered by exactly one fixture; reference examples report 0 errors
- [x]
RMI-THEORYOFCHANGESPEC-014Assumption, measurement, intervention, accountability, actor, evidence, pathway, metric, claim, budget, audience, cause, responsibility, risk, equity, validation, and provenance rules with seeded-defect fixtures - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-012,RMI-THEORYOFCHANGESPEC-036 - Satisfies: PRD-FR-303, PRD US-10, PRD US-11, TRD §6.4
- Acceptance: one fixture per rule; ≥ 15 rules total implemented and tested
- [x]
RMI-THEORYOFCHANGESPEC-015ToC spec template, core RubricSet (sector-neutral UNDG checklist plus plausible / feasible / testable), and development rubric profile for LLM-as-a-Judge, embedded inspec/ - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-004 - Satisfies: PRD-FR-601, TRD-012, TRD-028
- Acceptance: both rubrics parse into
rubric.RubricSet,Validate()passes, weights sum to 1.0; every UNDG checklist item maps to a criterion - [x]
RMI-THEORYOFCHANGESPEC-050pkg/judgeandimpact-model judge show | record | summary: bundle hashing, verdict validation, multi-judge agreement - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-015,RMI-THEORYOFCHANGESPEC-039 - Satisfies: PRD-FR-803, PRD G-16, PRD US-18, TRD-034
- Acceptance: bundle golden; verdicts with mismatched bundle hash rejected; aggregation reports per-category agreement
- [x]
RMI-THEORYOFCHANGESPEC-016impact-model lintCLI and rubric discrimination eval with at least two judges and agreement, recorded indocs/evals/ - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-013,RMI-THEORYOFCHANGESPEC-014,RMI-THEORYOFCHANGESPEC-050 - Satisfies: PRD US-2, PRD success metrics "LLM-judge discrimination" and "Judge agreement recorded"
- Acceptance: no judge raises a high or critical finding on the reference example, after which it goes to human review; seeded weak ToC fails ≥ 3 rubric categories; agreement reported
Phase 8 — Measurement¶
Theme: Turn spreadsheet data and observations into indicator, outcome, pathway, and assumption status with reproducible roll-ups and confidence — adopter milestone M2 Status: Completed — 6 of 6 items completed
- [x]
RMI-THEORYOFCHANGESPEC-017Indicator progress, expected progress, status (including baseline_only), and health mapping - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-008,RMI-THEORYOFCHANGESPEC-033 - Satisfies: PRD-FR-401, PRD-FR-402, PRD-FR-509, TRD-009, TRD-019
- Acceptance: table tests cover every status branch
- [x]
RMI-THEORYOFCHANGESPEC-018Outcome, pathway-status, and cross-pathway metric roll-ups through precondition logic and nested groups - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-009,RMI-THEORYOFCHANGESPEC-017 - Satisfies: PRD-FR-403, PRD-FR-113, TRD-019
- Acceptance: tests for all / any / at_least / nested groups and each metric aggregation
- [x]
RMI-THEORYOFCHANGESPEC-019Assumption propagation and link / path confidence scoring - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-018 - Satisfies: PRD-FR-404, PRD-FR-405
- Acceptance: invalidated assumption flags every downstream outcome (US-5); confidence tests with documented weights
- [x]
RMI-THEORYOFCHANGESPEC-020measure.Evaluatestatus snapshot, theme counts, andDiffbetween periods; determinism - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-019,RMI-THEORYOFCHANGESPEC-035 - Satisfies: PRD-FR-406, TRD INV-3
- Acceptance: golden snapshots byte-identical across two runs
- [x]
RMI-THEORYOFCHANGESPEC-040CSV observation import with mapping file and unmatched-row reporting - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-007,RMI-THEORYOFCHANGESPEC-033 - Satisfies: PRD-FR-204, PRD US-13, TRD-021
- Acceptance: seeded bad rows all reported; 0 silent drops
- [x]
RMI-THEORYOFCHANGESPEC-021impact-model statusandimpact-model importCLI; performance benchmarks - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-020,RMI-THEORYOFCHANGESPEC-040,RMI-THEORYOFCHANGESPEC-011 - Satisfies: PRD US-4, PRD US-13, TRD §8
- Acceptance: CLI tests; 200-node benchmark under 1 s
Phase 9 — Reporting, Story, and Presentations¶
Theme: Executive, audience story, learning-review, portfolio, narrative, and division-of-labour reports, plus exec / learning-review / story decks with optional export — adopter milestone M3 Status: Completed — 8 of 8 items completed
- [x]
RMI-THEORYOFCHANGESPEC-024Status reports: exec, results, assumptions, interventions, evidence-gaps, full - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-020,RMI-THEORYOFCHANGESPEC-022,RMI-THEORYOFCHANGESPEC-023 - Satisfies: PRD-FR-501, PRD US-3
- Acceptance: golden files; every reported status cites its inputs
- [x]
RMI-THEORYOFCHANGESPEC-041Audience story report with consent-gated quotes - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-024,RMI-THEORYOFCHANGESPEC-035 - Satisfies: PRD-FR-506, PRD US-14, TRD-023
- Acceptance: golden files; no-consent and restricted quotes absent
- [x]
RMI-THEORYOFCHANGESPEC-042Learning-review and portfolio reports - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-024,RMI-THEORYOFCHANGESPEC-034 - Satisfies: PRD-FR-507, PRD-FR-508, PRD US-14
- Acceptance: golden files over two observation periods
- [x]
RMI-THEORYOFCHANGESPEC-045Narrative (programme / strategy document scaffold) and division-of-labour reports; accountable vs monitored-only separation - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-022,RMI-THEORYOFCHANGESPEC-044 - Satisfies: PRD-FR-510, PRD-FR-511, PRD-FR-120, TRD-026, TRD-027
- Acceptance: golden files over a mixed-responsibility fixture
- [x]
RMI-THEORYOFCHANGESPEC-060Projection artifacts: outcome map (outcomes framework), logframe matrix, and logic-model diagram, as definition reports and in the definitions pack - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-022,RMI-THEORYOFCHANGESPEC-044 - Satisfies: PRD-FR-513
- Acceptance: golden files; every row names the element IDs it was projected from; a one-pathway theory and a multi-pathway theory both render
- [x]
RMI-THEORYOFCHANGESPEC-025Custom report templates and restricted-evidence suppression - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-024 - Satisfies: PRD-FR-502, PRD-FR-503, TRD-011
- Acceptance: custom template test; restricted excerpt absent from all outputs
- [x]
RMI-THEORYOFCHANGESPEC-051Exec, learning-review, and story decks;--export pdf|pptx|htmlvia optionalmmdcandmarp - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-049,RMI-THEORYOFCHANGESPEC-024,RMI-THEORYOFCHANGESPEC-041,RMI-THEORYOFCHANGESPEC-042 - Satisfies: PRD-FR-801, PRD-FR-802, PRD US-17, TRD-033
- Acceptance: golden decks; export test produces files when tools are present and exits 3 with guidance when absent
- [x]
RMI-THEORYOFCHANGESPEC-026impact-model reportstatus kinds in CLI and executive-report benchmark - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-025,RMI-THEORYOFCHANGESPEC-041,RMI-THEORYOFCHANGESPEC-042,RMI-THEORYOFCHANGESPEC-045,RMI-THEORYOFCHANGESPEC-021 - Satisfies: PRD G-5
- Acceptance: CLI tests; exec report for 200-node ToC under 1 s
Phase 10 — Agents and Revisions¶
Theme: MCP tools (including validated patching), agent skills, semantic theory diffs, and automated Google sharing — adopter milestone M4 Status: In progress — 3 of 4 items completed
- [x]
RMI-THEORYOFCHANGESPEC-027MCP server exposing schema, validate, lint, graph, status, report, patch, deck, and judge-bundle tools with root confinement - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-026,RMI-THEORYOFCHANGESPEC-046,RMI-THEORYOFCHANGESPEC-050 - Satisfies: PRD-FR-602, PRD-FR-706, PRD US-6, TRD-013
- Acceptance: in-memory transport test drives every tool; path traversal rejected;
impact_patchrejects error-introducing patches - [x]
RMI-THEORYOFCHANGESPEC-048pkg/diffsemantic theory diff andimpact-model diff --revision - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-046,RMI-THEORYOFCHANGESPEC-009 - Satisfies: PRD-FR-705, PRD US-16, PRD US-21, TRD-032
- Acceptance: golden diffs keyed by element ID, including rejected elements; draft Revision lists changed IDs
- [x]
RMI-THEORYOFCHANGESPEC-043Agent skills: impact-author (edits via patches), impact-status-refresh, impact-learning-review, impact-audience-story - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-026,RMI-THEORYOFCHANGESPEC-039,RMI-THEORYOFCHANGESPEC-047 - Satisfies: PRD-FR-605, PRD US-15, PRD G-13, TRD-024
- Acceptance: smoke test runs each skill's command sequence twice with identical outputs
- [ ]
RMI-THEORYOFCHANGESPEC-059toc-share-google agent skill (publish, read responses, import) over external Google tooling, with manual fallback - Type: capability · Status: planned
- Depends on:
RMI-THEORYOFCHANGESPEC-056,RMI-THEORYOFCHANGESPEC-057,RMI-THEORYOFCHANGESPEC-058,RMI-GOGOOGLE-006,RMI-MCPGOOGLE-004,RMI-MCPGOOGLE-005 - Satisfies: PRD-FR-607, PRD US-20, PRD G-18, TRD-042
- Acceptance: smoke test with recorded gogoogle and mcp-google outputs produces a valid ReviewResponse and proposals; manual fallback documented and tested via --from-csv
Phase 11 — Ecosystem Integration and Release¶
Theme: PRISM exports, the USM/UFM integration contract and competing-futures fixture, and the v0.1.0 release Status: In progress — 3 of 4 items completed
- [x]
RMI-THEORYOFCHANGESPEC-028pkg/prismxOKR and outcome-journey export with import-boundary test - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-020 - Satisfies: PRD-FR-603, PRD US-7, TRD-014
- Acceptance: exported OKR documents pass prism-roadmap validation; core packages do not import prism-roadmap
- [x]
RMI-THEORYOFCHANGESPEC-029Actor registry resolution, cross-document references, and competing-futures fixture - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-008,RMI-THEORYOFCHANGESPEC-032 - Satisfies: PRD G-8, PRD US-8, TRD §7.3
- Acceptance: three competing-future ToCs sharing one actor registry validate with no schema changes
- [ ]
RMI-THEORYOFCHANGESPEC-030USM/UFM integration contract document - Type: capability · Status: in_progress
- Depends on:
RMI-THEORYOFCHANGESPEC-029 - Satisfies: PRD G-8
- Acceptance: contract reviewed by USM/UFM owners
- [x]
RMI-THEORYOFCHANGESPEC-031Documentation, CHANGELOG, v0.1.0 release, and spec-type sync follow-up for specification-workflow-spec - Type: release · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-027,RMI-THEORYOFCHANGESPEC-043,RMI-THEORYOFCHANGESPEC-048,RMI-THEORYOFCHANGESPEC-059,RMI-THEORYOFCHANGESPEC-028,RMI-THEORYOFCHANGESPEC-030,RMI-THEORYOFCHANGESPEC-051 - Satisfies: all PRD success metrics (release gate)
- Acceptance: CI green before tag; publication scan clean; follow-up RMI filed in specification-workflow-spec
Phase 12 — Attribution and Notices¶
Theme: Credit the methods and projects the spec builds on, and carry the notices its dependencies require Status: Completed — 1 of 1 items completed
- [x]
RMI-THEORYOFCHANGESPEC-061Credits, third-party notices, and a dependency license audit - Type: capability · Status: completed
- Satisfies: — (documentation and compliance)
- Acceptance: README and docs credit the methods and projects used; THIRD_PARTY_NOTICES.md carries each distributed dependency's license and notice text; every module compiled into the CLI is permissively licensed
Phase 13 — Headline Metrics¶
Theme: Tie the causal chain to the metrics each program steers by and each audience is told: a North Star per program, guardrails, curated headline numbers per audience, and metrics worked out from others Status: Completed — 7 of 7 items completed
- [x]
RMI-THEORYOFCHANGESPEC-062Headline metrics per audience and a North Star per program, with a metric-chain diagram - Type: capability · Status: completed
- Satisfies: PRD G-8
-
Acceptance: The README and docs index show how resources, activities, outputs, outcomes, impact, North Star metrics, and audience headline metrics link; audiences can list an ordered, worded set of headline indicators; the story report and deck use it; lint checks it
-
[x]
RMI-THEORYOFCHANGESPEC-063Shared pathway and tier helpers used by every rule and report - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-062 -
Acceptance: One function says which pathways an indicator serves and which tier it is; PTH-002/003 and MET-001/003 agree; no rule re-derives membership
-
[x]
RMI-THEORYOFCHANGESPEC-064Model headline metrics: audience headline lists, pathway focus (North Star, one metric, guardrails), product kind, mission and vision - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-063 -
Acceptance: Go types, JSON Schemas regenerated, references validated; existing documents still validate
-
[x]
RMI-THEORYOFCHANGESPEC-065Lint the activity-to-impact chain, headline lists, and North Stars - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-064 -
Acceptance: Rules report an activity that leads nowhere, a link and intervention that disagree, an audience with no numbers, a North Star that is not an outcome at or below the ceiling
-
[x]
RMI-THEORYOFCHANGESPEC-066Reports: metric tree, headline use in story and deck, headline report, estimates shown as estimates - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-065 -
Acceptance: A metric-tree report per pathway; story and deck use the audience headline list and drop the hardcoded five; an estimated value reads as an estimate
-
[x]
RMI-THEORYOFCHANGESPEC-067Structured metric composition and derived indicator values - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-066 -
Acceptance: A metric can be a ratio or product of other metrics; a derived value is computed when its parts are observed; no cycles
-
[x]
RMI-THEORYOFCHANGESPEC-068Examples, guides, MCP text, and the metric-chain diagram updated for headline metrics - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-066,RMI-THEORYOFCHANGESPEC-067 - Acceptance: The Riverbend example uses the new fields; a guide explains tiers, North Star, and headline; the diagram's planned boxes become solid
Phase 14 — Vocabulary¶
Theme: Make the terms consistent and easy to learn Status: Completed — 1 of 1 items completed
- [x]
RMI-THEORYOFCHANGESPEC-069Rename the intermediate outcome level to medium_term so short, medium, and long form one series - Type: capability · Status: completed
- Acceptance: the level is medium_term in the model, schema, reports, examples, and docs; every test passes; no alias is kept
Phase 15 — Layered Logic Model Diagram¶
Theme: Draw the logic model as stage columns with real edges, from one shared diagram spec Status: Completed — 3 of 3 items completed
- [x]
RMI-THEORYOFCHANGESPEC-070Derive a layered logic model graph (stages, nodes, edges) from a view - Type: capability · Status: completed
-
Acceptance: inputs, activities, outputs, and outcomes by level are columns; edges come from resources, interventions, and links; order is deterministic and reduces crossings; tests cover each edge kind
-
[x]
RMI-THEORYOFCHANGESPEC-071Emit the logic model graph as D2 and Mermaid through d2vision and add report kinds - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-070 -
Acceptance: logic-model-layered and logic-model-d2 reports; the existing logic-model diagram is unchanged; the D2 compiles and keeps every label; stage columns align
-
[x]
RMI-THEORYOFCHANGESPEC-072Use the layered logic model diagram in the workshop deck beside the existing slide - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-071 - Acceptance: the deck has the layered diagram as a slide and the existing table and diagram remain; docs describe both and how they differ
Phase 16 — Release v0.2.0¶
Theme: Prepare the second release: changelog, release notes, and guides Status: Completed — 1 of 1 items completed
- [x]
RMI-THEORYOFCHANGESPEC-073Prepare v0.2.0: changelog, release notes, and guide updates - Type: capability · Status: completed
- Acceptance: CHANGELOG.json has v0.2.0 and CHANGELOG.md is generated from it; docs/releases/v0.2.0.md exists; the guides cover headline metrics, composed indicators, and the layered diagram; mkdocs build --strict passes
Phase 17 — Rename to impact-model¶
Theme: One impact model, with the theory of change, logic model, and other methodologies as projections of it Status: Completed — 4 of 4 items completed
- [x]
RMI-THEORYOFCHANGESPEC-074Rename the module, packages, command, kind, schemas, MCP tools, and skills to impact-model - Type: capability · Status: completed
-
Acceptance: module path github.com/ProductBuildersHQ/impact-model; pkg/impact and pkg/impactio; command impact-model; kind ImpactModel; schemas impact-model; tools impact_; skills impact-*; build, vet, tests, and lint pass
-
[x]
RMI-THEORYOFCHANGESPEC-075Rewrite the documentation around one impact model with several views - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-074 -
Acceptance: README, docs index, concepts, guides, specs, and examples use the impact-model names, commands, and file names; the theory of change and the logic model are described as views of one model and not one of the other
-
[x]
RMI-THEORYOFCHANGESPEC-076Record the rename in an architecture decision record, the changelog, and the roadmap - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-075 -
Acceptance: an ADR states the decision, what changed, and what did not; CHANGELOG.json notes the breaking rename; the roadmap lists the phase
-
[x]
RMI-IMPACTMODEL-002Present the impact model as a ProductBuildersHQ format whose projections include the theory of change, the logic model, OKRs, and V2MOM - Type: capability · Status: completed
- Depends on:
RMI-THEORYOFCHANGESPEC-076 - Acceptance: README and docs index lead with the impact model as the format and list its projections with how to get each; projection is the term for a method's rendering throughout the docs, distinct from report.View; planned projections are marked as planned
Phase 18 — Release v0.3.0¶
Theme: Prepare the first release as impact-model: changelog, release notes, and an upgrade guide for the rename Status: Completed — 1 of 1 items completed
- [x]
RMI-IMPACTMODEL-001Prepare v0.3.0: changelog, release notes, and the upgrade guide for the rename - Type: capability · Status: completed
- Acceptance: CHANGELOG.json has v0.3.0 and CHANGELOG.md is generated from it; docs/releases/v0.3.0.md maps every old name to its new one; README and docs index point to it; mkdocs build --strict passes
Phase 19 — Strategy Projections¶
Theme: Draw strategy and improvement formats such as V2MOM and DMAIC from the impact model, on prism-roadmap's goal types, linked by ID and kept as separate documents Status: In progress — 4 of 5 items completed
- [x]
RMI-IMPACTMODEL-003Export a V2MOM projection of an impact model - Type: capability · Status: completed
-
Acceptance: context.values added; export v2mom writes vision, values, methods, obstacles, and measures from the model, linked by ID; the V2MOM document stays separate
-
[x]
RMI-IMPACTMODEL-005Export a DMAIC projection of an impact model - Type: capability · Status: completed
-
Acceptance: export dmaic writes a prism-roadmap goals/dmaic document: categories from pathways, metrics from indicators with baseline, current, target, direction, status, and data points from observations, root causes from the problem tree, and initiatives from interventions; control limits and process capability only where observations support them, never invented
-
[x]
RMI-IMPACTMODEL-004Export OKRs through prism-roadmap's goals/okr instead of structured-goals - Type: chore · Status: completed
-
Acceptance: pkg/prismx imports prism-roadmap/goals/okr; structured-goals is gone from go.mod; the OKR export's golden output is unchanged
-
[x]
RMI-IMPACTMODEL-006Mark pathways for DMAIC and lint their measurement cadence and sample size - Type: capability · Status: completed
- Acceptance: pathway.projections accepts dmaic; lint checks that the indicators of a dmaic pathway declare a frequency and collection system, are measured at least 20 times a year, and survey indicators get at least 30 responses per measurement; thresholds match prism-roadmap's DMAIC code
-
Note: done except the threshold cross-check, which moves to
RMI-IMPACTMODEL-007; prism-roadmap does not export DMAIC thresholds yet, so they live inpkg/impact -
[ ]
RMI-IMPACTMODEL-007Cross-check the DMAIC thresholds with prism-roadmap once it exports them - Type: chore · Status: proposed
- Acceptance: prism-roadmap's goals/dmaic exports its minimum points and responses; a prismx test fails if impact.DMAICMinPoints or impact.DMAICMinResponsesPerPoint differ
Phase 20 — Release v0.4.0¶
Theme: Prepare the release with the V2MOM and DMAIC projections: changelog, release notes, and the upgrade note for the OKR types Status: Completed — 1 of 1 items completed
- [x]
RMI-IMPACTMODEL-008Prepare v0.4.0: changelog, release notes, and the upgrade note for ToOKR - Type: capability · Status: completed
- Acceptance: CHANGELOG.json has v0.4.0 and CHANGELOG.md is generated from it; docs/releases/v0.4.0.md covers V2MOM, DMAIC, the model additions, and the Go upgrade for prismx.ToOKR; mkdocs build --strict passes
Phase 21 — Problem Trees¶
Theme: Causes as a network, effects as branches, and a problem-tree projection drawn from a JSON diagram spec, with arrows flowing up from the roots to the core problem and on to the branches Status: Completed — 4 of 4 items completed
- [x]
RMI-IMPACTMODEL-009Model effects and a network of causes - Type: capability · Status: completed
- Satisfies: PRD-FR-119, PRD-FR-126, TRD-025
-
Acceptance: causes take parents[] and provenance; effects with statement, parents[], affected groups, evidence, out of scope, and provenance; outcomes take addressesEffects; validation per TRD 7.1 (VAL-CAU-003/004, VAL-EFF-001..004); schema regenerated; the edit engine and element commands handle effects and parents
-
[x]
RMI-IMPACTMODEL-010Lint effects that nothing relieves - Type: capability · Status: completed
- Depends on:
RMI-IMPACTMODEL-009 - Satisfies: PRD-FR-305
-
Acceptance: IMP-EFF-001 and IMP-EFF-002 per TRD 6.4, with seeded-defect tests
-
[x]
RMI-IMPACTMODEL-011Draw the problem tree from a JSON diagram spec - Type: capability · Status: completed
- Depends on:
RMI-IMPACTMODEL-009 - Satisfies: PRD-FR-514, TRD-043
-
Acceptance: pkg/problemtree layers causes below and effects above the core problem per TRD 6.6, exports a d2vision diagram spec, and renders D2 and Mermaid from it; reports problem-tree, problem-tree-d2, and problem-tree-json; golden tests
-
[x]
RMI-IMPACTMODEL-012Document problem trees and give the examples effects - Type: capability · Status: completed
- Depends on:
RMI-IMPACTMODEL-010,RMI-IMPACTMODEL-011 - Acceptance: a problem tree guide with the mapping to the diagram; the examples have effects and outcomes that address them; README, concepts, CLI reference, changelog, and roadmap updated
Phase 22 — Release v0.5.0¶
Theme: Prepare the release with problem trees: changelog, release notes, guide updates, and the migration from parent to parents Status: Completed — 1 of 1 items completed
- [x]
RMI-IMPACTMODEL-013Prepare v0.5.0: changelog, release notes, guide updates, and the parent to parents migration - Type: capability · Status: completed
- Acceptance: CHANGELOG.json has v0.5.0 and CHANGELOG.md is generated from it; docs/releases/v0.5.0.md covers problem trees and the migration from parent to parents; the user guides cover effects; mkdocs build --strict passes
Phase 23 — Exports for Agents and an Overview Refresh¶
Theme: The PRISM exports through the service layer and the MCP server, the problem tree in the definitions pack, and an overview, credits, and guides that cover every projection Status: Completed — 3 of 3 items completed
- [x]
RMI-IMPACTMODEL-014Export through the service layer and from the MCP server - Type: capability · Status: completed
-
Acceptance: service.ExportPRISM draws the OKR, journey, V2MOM, and DMAIC exports for the command line and the MCP server; the impact_export tool takes observations and a status date and refuses a model the format cannot be drawn from; tests for both
-
[x]
RMI-IMPACTMODEL-015Include the problem tree in the definitions pack - Type: capability · Status: completed
-
Acceptance: definitions pack opens with the problem tree when the model records causes or effects; goldens updated
-
[x]
RMI-IMPACTMODEL-016Refresh the overview, credits, and guides for V2MOM, DMAIC, and problem trees - Type: documentation · Status: completed
- Acceptance: README and docs/index.md list every projection and export; credits name V2MOM, DMAIC, and problem-tree sources; a test keeps README and docs/index.md in step