0 Fork 0

mirror of https://github.com/paperclipai/paperclip.git synced 2026-10-27 19:35:13 +01:00

Files

T

Co-Authored-By: Paperclip <noreply@paperclip.ing>

2026-08-11 12:56:59 +00:00

Native Runner Knowledge Update Log

2026-08-07

Creation: Established the OKF v0.2 bundle, typed Conformance journal, verification evidence, template, and progressive-disclosure indexes.
Rust boundary correction: Added the production-direction Cargo workspace, shared-fixture parity evidence, negative Cargo boundary coverage, and a superseding journal entry after the Conformance checkpoint exposed the TypeScript-only omission.
Replay static replay: Added executable PRP v1 schemas, cross-language fixture/golden parity, deterministic reducer, CLI/browser replay, compatibility policy, tutorial, screenshot, and verification evidence.
Local runner: Added the deterministic Rust supervisor, stdio fake-harness protocol, scripted scenarios, bounded logs, cleanup and terminal guards, TypeScript mock core, validated live browser mode, screenshots, tutorial, and verification evidence.
Durable recovery: Added package-local WebSocket authentication, a durable outbox, cumulative acknowledgements, restart replay, command deduplication, recovery diagnostics, tutorial, and evidence.

Codex skillless Codex driver: Added the direct app-server v2 driver, compact task envelope, isolated model context, semantic completion tools, canonical event/replay mapping, safe real-model example, tutorial, journal, and exact verification trace.
Live console UX gate: Approved the live-console interaction map, the shadcn/AI Elements adapt/reject decision record with a zero-new-runtime- dependency rule, keyboard/accessibility acceptance criteria, four token additions, and annotated desktop/mobile mockups.
Live console protocol/server layer: Added deterministic Codex request, goal, lineage, control-race, reconnect, replay, and redaction fixtures; a typed server-only demo boundary; and a passing real Codex server trace.
Live console: Added the package-local browser tracer over the canonical protocol — transcript, composer, same-turn steering, three interrupt races, runtime request cards, capability-gated goals, parent/child lineage, reconnect, refresh, replay, and a redacted protocol inspector — plus a deterministic demo driver, eleven demo chats, adapted shadcn/AI Elements components with no new runtime dependencies, and a boundary check that keeps it that way.
SDK browser SDK: Extracted versioned browser, React, and scoped-style exports; added a public-API-only reference console and mini consumer; proved the five approved extension points, keyboard/accessibility contracts, reconnect and replay parity, package boundaries, and safe real-Codex use.

Capability browser scenario explorer: Added the package-local scenario index with declared actor roles, task modes, and catalog claims; deterministic fixture seeds and fake-agent plans for all 106 eval cases; a run-artifact runtime producing tool exposure, control-plane actions, authorization records, a ten-domain immutable state diff, and a parity verdict that carries Capability output through without re-judging it; the read-only explorer over the frozen 0.1.2 SDK with hash routing, both transcript channels, and four inspector tabs; 45 unit and 19 browser tests; and the twenty-four-image screenshot acceptance set.
Standalone integration design: Specified the default-off thin Paperclip adapter seam, public port reconciliation, first-class persisted mode/result recovery, canonical per-run event allocation, typed credential-free native input, complete Section 18 finalization/arbitration, kill switch, exact file allowlist, conformance matrix, and mock/native/legacy tutorial commands before production code.
Standalone integration implementation: Added persisted default-off runtime selection, the database-backed public control-plane port, package-owned normalized Codex sessions, canonical event allocation/replay, immutable completion/result/finalization records, server-owned status arbitration and audit, native cancellation, migration coverage, a runnable tutorial, and deterministic mock/real-port evidence while retaining legacy as the kill- switch path.

Standalone scope correction: Superseded live-instance installation and rollout work. Added a package-local native/legacy simulator, standalone demo page, default-legacy selection, kill-switch rollback, shared public-port conformance, reducer/replay/finalization inspection, and a no-instance tutorial under packages/paperclip-runner/.
Capability clean-room live chat: Added a second primary path beside the preset scenario explorer: a blank #/chat surface that mints a fresh mock tenant (company, agent, one blank issue), drives real Codex through real runnerd with a deterministic exposure profile, keeps tool, policy, event, and state evidence in an on-demand drawer, rotates session authority and mock identities on reset and new chat, and publishes the real-API block as an inspectable control-plane record.