Update roadmap from experiment results

[tonyketcham]

Source: flatbread-flow-pmf-audit.md; updated with flatbread-agent-artifact-opportunity.md

Project metadata

• Phase: Experiments and adoption validation
• Sequence: 28
• Priority: P2
• Suggested owner role: product
• Supporting roles: none
• Dependency status: blocked
• Blocked by: Benchmark canonical starter path to first typed read result #162 Run relational starter benchmark; Run TypeScript safety interview/test #163 Run TypeScript safety interview/test; Run export trust self-review demo #164 Run export trust experiment; Document representative Effort Graph reference-layout mapping #167 Run Effort Graph wire-up against real agent artifact folders; Document three-layout Effort Graph schema mapping #168 Run adversarial Effort Graph schema test across harness layouts; Benchmark representative artifact-byte reduction for blocking-decision retrieval #169 Benchmark cold-start vs Flatbread-mediated agent artifact retrieval

Task details

• Revisit the PMF improvement roadmap after the core adoption experiments and the agent artifact validation track both complete.
• Turn evidence into explicit keep, kill, or iterate decisions for the next product slice.
• State whether the agent artifact / Effort Graph opportunity should become a primary wedge, remain a secondary vertical, or be deferred.

Acceptance criteria

• Roadmap priorities are updated based on benchmark/interview/export and agent artifact validation evidence.
• Each major initiative has a keep, kill, or iterate decision with rationale.
• The output explicitly decides the status of the agent artifact / Effort Graph opportunity.
• Follow-up issues are created or closed to reflect the updated priorities.

[tonyketcham]

This follow-up is based on changes in PR #177.

• Landed: docs/roadmap.md reflects the experiment results with updated priorities and keep/kill/iterate decisions (including the Effort Graph status call).
• Still remaining: the acceptance step depends on maintainer-side GitHub hygiene—creating/closing the follow-up issues that the roadmap drafts.

Decision: confirm whether #165 should stay open until that operational follow-through is done, or be closed now with explicit manual ownership for those follow-up board actions.