Evaluate repository briefings and podcast derivatives from canonical content #349

Outcome

Evaluate and prototype a reusable Renderflow profile that can turn approved repository material into an accessible audio briefing or podcast-like derivative without creating a competing source of project truth.

Reading List is the initial experience reference: https://www.readinglist.live/.

Candidate inputs

• README and getting-started documentation;
• released roadmap/current-state summaries;
• architecture overview/booklet specifications;
• changelogs and release notes;
• research/publication abstracts;
• explicitly approved narration notes.

Architecture

canonical reviewed content
        ↓
bounded episode/briefing plan
        ↓
reviewable narration script
        ↓
optional TTS/voice adapter
        ↓
audio + transcript + chapters + feed metadata

Renderflow owns deterministic planning, scripts, manifests, packaging, and provider-neutral transforms. A voice provider is optional. Product repositories approve content and publication. Identity supplies approved audio/voice guidance when available.

Scope

• Compare short repository briefing, release-update feed, narrated documentation, and true podcast formats.
• Define episode identity, source revision, freshness, chapters, links, transcript, accessibility, license, voice provenance, and approval state.
• Support human narration and optional TTS without making synthetic audio mandatory.
• Produce accessible transcript/HTML and an RSS-compatible projection when justified.
• Prevent stale generated audio from appearing current after source changes.
• Pilot one short synthetic/repository-safe briefing.

Acceptance criteria

• The research recommends adopt, adapt, defer, or reject for each format.
• Audio never becomes the only way to access required information.
• Every episode points to exact reviewed source revisions.
• Script and publication approval are explicit.
• Voice/model/provider use, license, and provenance are recorded.
• The core workflow works locally without a hosted provider.
• Transcript, chapters, metadata, and feed output validate.
• Stale or superseded episodes remain clearly labeled.
• No repository publishes a feed merely to satisfy a template.

Non-goals

• Fabricating conversations, hosts, guests, or endorsements.
• Cloning Reading List.
• Automatically publishing unreviewed synthetic voices.
• Using a person's voice without explicit permission.

Audit-to-roadmap reconciliation — 2026-10-21

Audit: https://github.com/egohygiene/renderflow/blob/audit/flow-suite-20261021T025401Z/audits/flow-suite-20261021T025401Z.md · Complete roadmap: #367

Evaluate adopt/adapt/defer/reject first; an audio branch never becomes the only access to required source information.

Bounded preparation: Review existing approved-content profile contracts; draft a bounded evaluation and source/approval model.

Remaining acceptance (common qualification deferral; no dependency star): Chosen provider/assets, approved content and executable qualification for any adopted path.

The maintainer's implementation-sprint qualification deferral remains in force. This audit ran no Renderflow tests, compiler/build/Cargo resolution, product lint/format/schema qualification, native/CLI/provider/visual execution, installers or new CI. Reading existing exact-source provider records is not a qualification rerun. Source implementation and authored tests are not passing-test, native-fidelity, release or downstream-consumer evidence.

This is an additive planning note. The full original issue, acceptance criteria, scope, parent references and historical checkpoints remain authoritative and unchanged. No acceptance box or issue is closed by this audit. Recommended ledger parent: #367.