docs: align AgentFlow 1.0 narrative

[smota]

Implemented issues

• Implements feat: intelligent collaboration for AgentFlow SDLC epic #115

Related issues

• Refs feat: define intelligent collaboration modes for AgentFlow SDLC #109
• Refs feat: package AgentFlow SDLC roles as reusable role agents #110
• Refs feat: add harness collaboration plan resolver #111
• Refs feat: validate collaboration and sub-agent evidence #112
• Refs feat: implement controlled harness collaboration patterns #113
• Refs feat: add intelligent collaboration examples and evals #114

Workflow evidence

• Workflow-status comment: exception:copy/narrative docs follow-up requested directly after feat: intelligent collaboration for AgentFlow SDLC epic #115 merge; no separate issue workflow comment created
• Handover comments: exception:single-session docs narrative follow-up, no role transition handover needed
• Role-pass summary: Updated repository-facing narrative to align with AgentFlow SDLC 1.0 website positioning: use AI more confidently, first mature release, intelligent collaboration as value-first, smallest sufficient collaboration, parent synthesis, compact evidence.
• Capability evidence: not-applicable:documentation-only narrative update
• Collaboration evidence: single-agent:documentation-only update did not need helper agents
• Validation evidence: node scripts/verify-agent-workflow.mjs, pnpm format:check, and pnpm test passed.

CI-equivalent validation

• Status: passed
• Commands: node scripts/verify-agent-workflow.mjs; pnpm format:check; pnpm test.
• Notes: none

Release decision

• Required for release/promotion PRs; use not-applicable:<reason> for ordinary implementation PRs.
• Intended version/tag: not-applicable:documentation PR only
• Bump type: not-applicable:documentation PR only
• Rationale: aligns docs with upcoming 1.0 narrative; does not publish release itself
• Included issues: feat: intelligent collaboration for AgentFlow SDLC epic #115
• Deferred issues: none
• Release notes: docs/releases/v1.0.0.md
• Post-merge closeout: not-applicable:documentation PR only
• Closeout verification: not-applicable:documentation PR only
• Approval: human/operator merge by default

Role attribution matrix

Agent review

• Implemented by: pi
• Launcher: pi
• Executor: pi-parent
• Transport: local-cli
• Delegation boundary: current-session
• Model / runtime: not recorded
• Review: self-review
• Workflow profile: bounded
• Merge owner: human/operator
• Fallback chain: none
• Regression test: not-applicable:documentation-only update
• Mode: single-agent
• Self-review disclosure: not-applicable

Follow-up issues

• none