Merge pull request #702 from AssetOverflow/feat/wb-journal
feat(workbench): turn evidence journal + trace API
This commit is contained in:
Shay
commit
6edf2d3ac1
7 changed files with 848 additions and 5 deletions
4
.gitignore
vendored
4
.gitignore
vendored
|
|
@ -8,6 +8,10 @@ __pycache__/
|
||||||
traces/
|
traces/
|
||||||
.formation_cache/
|
.formation_cache/
|
||||||
|
|
||||||
|
workbench_data/*
|
||||||
|
!workbench_data/
|
||||||
|
!workbench_data/README.md
|
||||||
|
|
||||||
core-rs/target/
|
core-rs/target/
|
||||||
core-rs/Cargo.lock
|
core-rs/Cargo.lock
|
||||||
|
|
||||||
|
|
|
||||||
391
docs/workbench/wave-1-evidence-spine.md
Normal file
391
docs/workbench/wave-1-evidence-spine.md
Normal file
|
|
@ -0,0 +1,391 @@
|
||||||
|
# Wave 1 — Evidence Spine
|
||||||
|
|
||||||
|
Status: approved plan
|
||||||
|
Date: 2026-06-12
|
||||||
|
Supersedes: implementation-plan.md phases W-032+
|
||||||
|
Peer-reviewed by: Codex (architectural path selection), GPT5.5-Thinking (trace
|
||||||
|
honesty correction), Claude Opus 4.6 (original plan + synthesis)
|
||||||
|
|
||||||
|
## Governing idea
|
||||||
|
|
||||||
|
Evidence context is the intrinsic UI space. Every route is a projection of the
|
||||||
|
same evidence manifold. The workbench is not organized around routes (boxes on
|
||||||
|
screen); it is organized around the evidence chain:
|
||||||
|
|
||||||
|
```
|
||||||
|
operator intent
|
||||||
|
-> selected evidence subject
|
||||||
|
-> provenance
|
||||||
|
-> admissibility
|
||||||
|
-> replay
|
||||||
|
-> authority
|
||||||
|
-> allowed action
|
||||||
|
```
|
||||||
|
|
||||||
|
Chat shows the newest turn. Trace deepens it. Replay tests it. Proposals ask
|
||||||
|
whether it may alter reviewed memory. Packs/Vault/Audit reveal the substrate
|
||||||
|
that made it possible. Once the spine exists, remaining routes become inevitable
|
||||||
|
projections — not independent features.
|
||||||
|
|
||||||
|
## Three governing principles (ADR-0160)
|
||||||
|
|
||||||
|
1. **Audit-native, not analytics theater.** Every panel answers: what happened,
|
||||||
|
why was it allowed, what evidence exists, can it replay, who holds authority.
|
||||||
|
2. **Calm default, infinite depth.** Quiet surface by default; the deeper you
|
||||||
|
inspect, the more transparent it becomes.
|
||||||
|
3. **Replay before persuasion.** Show deterministic evidence before asking the
|
||||||
|
operator to trust anything.
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## Wave 1 deliverables
|
||||||
|
|
||||||
|
Six pieces, one architectural idea.
|
||||||
|
|
||||||
|
### 1A. Command Registry + Full Navigation
|
||||||
|
|
||||||
|
- [ ] Replace hardcoded command list in `CommandPalette.tsx` with a
|
||||||
|
route-registered command registry
|
||||||
|
- [ ] Each route registers its commands on mount via a shared context/provider
|
||||||
|
- [ ] Fuzzy search across routes, recent resources (turns, proposals, artifacts),
|
||||||
|
and actions (run eval, copy hash)
|
||||||
|
- [ ] `Cmd+K` opens, type-ahead filters, arrow keys navigate, `Enter` executes,
|
||||||
|
`Esc` closes
|
||||||
|
- [ ] Recent items: last 10 visited resources (turns, proposals, artifacts)
|
||||||
|
- [ ] Test: palette finds and navigates to every route and registered command
|
||||||
|
|
||||||
|
**Current state:** `CommandPalette.tsx` has only Chat/Proposals/Evals hardcoded.
|
||||||
|
|
||||||
|
**Key files:**
|
||||||
|
- `workbench-ui/src/design/components/primitives/CommandPalette.tsx`
|
||||||
|
- New: `workbench-ui/src/app/commandRegistry.ts` (or context provider)
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
### 1B. RightInspector as Evidence Drawer
|
||||||
|
|
||||||
|
- [ ] Remove permanent `collapsed=true` from `Shell.tsx`
|
||||||
|
- [ ] Create shared evidence-subject context: `useEvidenceSubject()` hook +
|
||||||
|
provider in Shell
|
||||||
|
- [ ] Each route pushes its selection (turn, proposal, artifact, pack, eval
|
||||||
|
result) into the shared context
|
||||||
|
- [ ] Inspector renders the appropriate evidence projection for the selected
|
||||||
|
subject type
|
||||||
|
- [ ] Toggle via `Cmd+I` or "inspect" affordance on selectable items
|
||||||
|
- [ ] Stays open across route transitions (operator opened it deliberately)
|
||||||
|
- [ ] Collapsed by default on fresh load (calm default)
|
||||||
|
- [ ] Resizable width via drag handle
|
||||||
|
- [ ] Test: inspector opens, shows correct context for Chat selection, closes,
|
||||||
|
persists across route change
|
||||||
|
|
||||||
|
**Current state:** `RightInspector.tsx` returns null. `Shell.tsx` hardcodes
|
||||||
|
`collapsed={true}`.
|
||||||
|
|
||||||
|
**Key files:**
|
||||||
|
- `workbench-ui/src/app/RightInspector.tsx`
|
||||||
|
- `workbench-ui/src/app/Shell.tsx`
|
||||||
|
- New: `workbench-ui/src/app/evidenceContext.ts`
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
### 1C. Minimal Evidence Primitives
|
||||||
|
|
||||||
|
Build only the six components the spine needs. Defer everything else until a
|
||||||
|
route proves it requires the component.
|
||||||
|
|
||||||
|
| Component | Purpose | Used by |
|
||||||
|
|---|---|---|
|
||||||
|
| `SplitPane` | Resizable horizontal/vertical split for list-detail | Trace, Proposals, Evals, Replay |
|
||||||
|
| `TabBar` | Accessible tab switching (see dependency note below) | Trace evidence sections, Inspector |
|
||||||
|
| `MetadataTable` | Key-value pair display for structured metadata | Trace, Proposals, Artifacts |
|
||||||
|
| `DigestBadge` | Copyable hash/digest with truncation + verify indicator | Trace, Replay, everywhere |
|
||||||
|
| `Timestamp` | Relative + absolute time, timezone-aware (PST/PDT) | All list views, Inspector |
|
||||||
|
| `SearchInput` | Filtered search with keyboard shortcut binding (`/`) | Command palette, list views |
|
||||||
|
|
||||||
|
For each component:
|
||||||
|
|
||||||
|
- [ ] Built with design tokens only (no raw hex/rgb)
|
||||||
|
- [ ] Motion via `--motion-duration-*` and `--motion-ease-*` tokens
|
||||||
|
- [ ] `prefers-reduced-motion` collapses to instant
|
||||||
|
- [ ] `:focus-visible` ring via `--color-focus-ring`
|
||||||
|
- [ ] Renders in PreviewPage (`/preview`)
|
||||||
|
- [ ] Unit test
|
||||||
|
|
||||||
|
**TabBar dependency note:** `@radix-ui/react-tabs` is not currently in
|
||||||
|
`package.json` (only `react-dialog` and `react-popover` are). Either add the
|
||||||
|
dependency explicitly with lockfile update, or implement TabBar with native
|
||||||
|
ARIA tab semantics without Radix. Decide at implementation time.
|
||||||
|
|
||||||
|
**Deferred primitives** (build when a route needs them):
|
||||||
|
- DataTable, TreeView, Timeline, CodeViewer, Drawer, Toast, SkeletonLoader, Kbd
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
### 1D. Workbench Turn Evidence Journal
|
||||||
|
|
||||||
|
**Critical correction (GPT5.5 review):** The workbench backend does NOT
|
||||||
|
currently attach a telemetry sink to chat turns. `WorkbenchApi()` constructs
|
||||||
|
with no sink, `_run_chat_turn()` creates a bare `ChatRuntime()`, and
|
||||||
|
`serialize_turn_event` redacts content by default. Raw runtime telemetry does
|
||||||
|
not contain the three surfaces needed for Trace.
|
||||||
|
|
||||||
|
**Solution:** A Workbench Turn Evidence Journal — a local, append-only,
|
||||||
|
content-bearing record of the `ChatTurnResult` envelope already returned by
|
||||||
|
`/chat/turn`.
|
||||||
|
|
||||||
|
This is a **read model**, not a cognitive runtime fork. It does not replace
|
||||||
|
runtime telemetry; it records the exact evidence the operator already saw.
|
||||||
|
|
||||||
|
#### Backend
|
||||||
|
|
||||||
|
- [x] New module: `workbench/journal.py`
|
||||||
|
- [x] `TurnJournal` class: append-only JSONL writer
|
||||||
|
- [x] Each `/chat/turn` response is journaled with:
|
||||||
|
- `turn_id` (stable, sequential)
|
||||||
|
- `timestamp` (ISO-8601 UTC)
|
||||||
|
- `trace_hash` (from ChatTurnResult)
|
||||||
|
- `prompt` (content-bearing — explicit in schema)
|
||||||
|
- `surface`, `articulation_surface`, `walk_surface` (all three, kept
|
||||||
|
separate per api-contract-v1.md line 231)
|
||||||
|
- `grounding_source`, `epistemic_state`, `normative_clearance`
|
||||||
|
- `verdicts` (identity, safety, ethics)
|
||||||
|
- `refusal_emitted`, `hedge_injected`
|
||||||
|
- `proposal_candidates`
|
||||||
|
- `turn_cost_ms`
|
||||||
|
- `journal_digest` (SHA-256 of the serialized entry)
|
||||||
|
- [x] Journal path: `workbench_data/turn_journal.jsonl` (under repo root, not
|
||||||
|
under `engine_state/` or `teaching/`)
|
||||||
|
- [x] Content-bearing warning: journal entries contain user prompts and engine
|
||||||
|
surfaces. Document this in `workbench_data/README.md` (not as a text
|
||||||
|
header in the JSONL file — every line must be valid JSON)
|
||||||
|
- [x] Path confinement: journal writes only to `workbench_data/`
|
||||||
|
- [x] No journal writes to `teaching/`, `packs/`, `language_packs/data/`, or
|
||||||
|
`engine_state/`. Note: existing chat turns DO write `engine_state/`
|
||||||
|
through the normal runtime checkpoint path governed by ADR-0146/0150 —
|
||||||
|
that is existing behavior, not journal behavior.
|
||||||
|
|
||||||
|
#### New API endpoints
|
||||||
|
|
||||||
|
- [x] `GET /trace/turns` — list journal entries (summary: turn_id, timestamp,
|
||||||
|
prompt excerpt, surface excerpt, trace_hash, grounding_source)
|
||||||
|
- [x] `GET /trace/{turn_id}` — full journal entry for a turn (replaces the
|
||||||
|
current 404 behavior)
|
||||||
|
- [x] Pagination: `?limit=50&offset=0` (default limit 50)
|
||||||
|
- [x] Unknown turn_id returns 404 (not synthetic data)
|
||||||
|
|
||||||
|
#### Tests
|
||||||
|
|
||||||
|
- [x] Append-only behavior: entries are never modified or deleted
|
||||||
|
- [x] Stable ordering: entries are sequential by turn_id
|
||||||
|
- [x] Prompt/content size limits respected (max 4096 chars prompt)
|
||||||
|
- [x] Path confinement: journal cannot write outside `workbench_data/`
|
||||||
|
- [x] No journal writes to `teaching/`, `packs/`, `language_packs/data/`; no
|
||||||
|
NEW writes to `engine_state/` beyond existing chat checkpoint behavior
|
||||||
|
(ADR-0146/0150)
|
||||||
|
- [x] Journal digest is deterministic for identical content
|
||||||
|
- [x] Round-trip: `/chat/turn` response -> journal -> `/trace/{turn_id}` ->
|
||||||
|
identical evidence fields
|
||||||
|
|
||||||
|
#### Optional linkage to runtime telemetry
|
||||||
|
|
||||||
|
If `ChatRuntime` is later configured with a `JsonlFileSink`
|
||||||
|
(`include_content=True`), the Trace route can cross-reference journal entries
|
||||||
|
with runtime telemetry events by `trace_hash`. This is additive — the journal
|
||||||
|
is the primary read model.
|
||||||
|
|
||||||
|
#### Public interfaces and types
|
||||||
|
|
||||||
|
Backend (Python):
|
||||||
|
- `workbench/journal.py` — `TurnJournalEntry` dataclass, `TurnJournalSummary`
|
||||||
|
dataclass, `TurnJournal` class (append, list, get)
|
||||||
|
- `workbench/schemas.py` — add `TurnJournalEntrySchema`, `TurnJournalSummarySchema`
|
||||||
|
for API serialization
|
||||||
|
|
||||||
|
Frontend (TypeScript):
|
||||||
|
- `workbench-ui/src/types/api.ts` — add `TurnJournalEntry`,
|
||||||
|
`TurnJournalSummary` interfaces mirroring the Python shapes
|
||||||
|
- `workbench-ui/src/api/client.ts` — add `fetchTraceTurns(limit?, offset?)`,
|
||||||
|
`fetchTraceTurn(turnId)`
|
||||||
|
- `workbench-ui/src/api/queries.ts` — add `useTraceTurns()`,
|
||||||
|
`useTraceTurn(turnId)` React Query hooks
|
||||||
|
- `workbench-ui/src/app/evidenceContext.ts` — `EvidenceSubject` union type:
|
||||||
|
`{ kind: 'turn', data: TurnJournalEntry }` | `{ kind: 'proposal', ... }` |
|
||||||
|
`{ kind: 'artifact', ... }` | `{ kind: 'eval_result', ... }` |
|
||||||
|
`{ kind: 'none' }`
|
||||||
|
|
||||||
|
All new API responses use the existing `{ ok, generated_at, data/error }`
|
||||||
|
envelope.
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
### 1E. Trace Route
|
||||||
|
|
||||||
|
- [ ] Replace `TraceRoutePlaceholder.tsx` with real Trace route
|
||||||
|
- [ ] Layout: `SplitPane` — turn timeline (left) + trace evidence panel (right)
|
||||||
|
- [ ] Turn timeline: list of journal entries with `DigestBadge` (trace_hash
|
||||||
|
thumbnail), `Timestamp`, prompt excerpt
|
||||||
|
- [ ] Trace evidence panel: `TabBar` with sections:
|
||||||
|
- **Surfaces** — all three, labeled explicitly (surface = user response,
|
||||||
|
walk_surface = telemetry evidence, articulation_surface = realizer
|
||||||
|
output). This IS the canonical proof of the api-contract-v1.md surface
|
||||||
|
separation contract.
|
||||||
|
- **Grounding** — source, epistemic state, normative clearance
|
||||||
|
- **Verdicts** — identity, safety, ethics verdicts with badge indicators
|
||||||
|
- **Metadata** — `MetadataTable` showing turn_cost_ms, checkpoint_emitted,
|
||||||
|
refusal/hedge status, proposal candidates
|
||||||
|
- **Raw** — collapsed-by-default full JSON viewer (StableJsonViewer)
|
||||||
|
- [ ] Selection pushes turn into evidence-subject context (RightInspector shows
|
||||||
|
same turn from inspector angle)
|
||||||
|
- [ ] `SearchInput` for filtering turns by prompt text or trace_hash prefix
|
||||||
|
- [ ] Empty state when journal is empty: "No turns recorded yet. Use Chat to
|
||||||
|
create evidence."
|
||||||
|
- [ ] Test: navigate to Trace, see real journal entries, select one, see evidence
|
||||||
|
panel, inspect raw JSON
|
||||||
|
|
||||||
|
**Key design rules:**
|
||||||
|
- Versor condition (when available): green < 1e-6, red >= 1e-6
|
||||||
|
- Walk surface labeled "telemetry/evidence" — never confused with user surface
|
||||||
|
- Trace hash always visible and copyable (replay before persuasion)
|
||||||
|
- Raw trace behind explicit expand (calm default, infinite depth)
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
### 1F. Mutation Doctrine Reconciliation
|
||||||
|
|
||||||
|
- [ ] Update `docs/workbench/implementation-plan.md` mutation section to match
|
||||||
|
reality
|
||||||
|
- [ ] Update `docs/workbench/acceptance-gates.md` to reflect admitted corridors
|
||||||
|
- [ ] Document the honest rule:
|
||||||
|
|
||||||
|
**The mutation rule is not "no buttons ever." It is:**
|
||||||
|
|
||||||
|
1. **Admitted corridor** — mutation only through an ADR-governed path
|
||||||
|
(math ratification via ADR-0172, chat turns via ADR-0146/0150).
|
||||||
|
2. **Explicit preconditions** — the UI shows what must be true before
|
||||||
|
mutation is allowed.
|
||||||
|
3. **Telemetry** — every mutation emits an auditable event.
|
||||||
|
4. **Replay evidence** — the operator can see replay-equivalence status
|
||||||
|
before acting.
|
||||||
|
|
||||||
|
`RatificationCommandPanel.tsx` already implements this pattern for math
|
||||||
|
proposals. This is the template for future mutation surfaces, not an exception
|
||||||
|
to a "no mutation" rule.
|
||||||
|
|
||||||
|
- [ ] Record what already exists: `ratify_math_proposal`, `reject`, `defer`
|
||||||
|
in `workbench/api.py` lines 112+; `RatificationCommandPanel.tsx` with
|
||||||
|
precondition gates
|
||||||
|
- [ ] No new mutation endpoints in Wave 1 beyond what exists
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## Wave 2 — Projections (after spine is live)
|
||||||
|
|
||||||
|
Once the evidence spine exists, each remaining route becomes a projection.
|
||||||
|
These are parallelizable.
|
||||||
|
|
||||||
|
### Packs Route
|
||||||
|
|
||||||
|
- [ ] Backend: `GET /packs`, `GET /packs/{pack_id}`
|
||||||
|
- [ ] Frontend: pack list with verification badges, lexicon browser
|
||||||
|
- [ ] New primitive: `TreeView` (pack hierarchy)
|
||||||
|
- [ ] Pushes selected pack into evidence-subject context
|
||||||
|
|
||||||
|
### Vault Route
|
||||||
|
|
||||||
|
- [ ] Backend: `GET /vault/summary`, `GET /vault/entries`,
|
||||||
|
`GET /vault/entries/{entry_id}`
|
||||||
|
- [ ] Frontend: entry list with epistemic state badges, recall history
|
||||||
|
- [ ] Pushes selected entry into evidence-subject context
|
||||||
|
|
||||||
|
### Audit Route
|
||||||
|
|
||||||
|
- [ ] Backend: `GET /audit/events`, `GET /audit/events/{event_id}`
|
||||||
|
- [ ] Frontend: vertical event timeline, mutation boundary highlighting
|
||||||
|
- [ ] New primitive: `Timeline`
|
||||||
|
- [ ] Pushes selected event into evidence-subject context
|
||||||
|
|
||||||
|
### Runs Route
|
||||||
|
|
||||||
|
- [ ] Backend: `GET /runs`, `GET /runs/{session_id}`
|
||||||
|
- [ ] Frontend: session list with checkpoint badges, turn history
|
||||||
|
- [ ] Cross-links to Trace for any turn
|
||||||
|
- [ ] Pushes selected session into evidence-subject context
|
||||||
|
|
||||||
|
### Settings Route
|
||||||
|
|
||||||
|
- [ ] Frontend (mostly localStorage): inspector default, JSON depth, timestamp
|
||||||
|
format, API connection
|
||||||
|
- [ ] Runtime config display (read-only)
|
||||||
|
- [ ] No dangerous mutations — engine config changes require CLI
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## Wave 3 — Polish + Demo Theater
|
||||||
|
|
||||||
|
### Existing module polish
|
||||||
|
|
||||||
|
- [ ] Chat: multi-line composer, submission history, richer evidence strip
|
||||||
|
- [ ] Proposals: visual chain diagram, provenance links, metric deltas
|
||||||
|
- [ ] Eval Center: failure-first display, lane health overview, run progress
|
||||||
|
- [ ] Replay Theater: synchronized side-by-side diff, multi-artifact comparison
|
||||||
|
|
||||||
|
### Demo Theater route
|
||||||
|
|
||||||
|
- [ ] Backend: `GET /demos`, `POST /demos/{demo_id}/run`,
|
||||||
|
`GET /demos/{demo_id}/scenarios`
|
||||||
|
- [ ] Frontend: demo list, scenario results with evidence, "what this proves" /
|
||||||
|
"what this does not prove" honesty cards
|
||||||
|
- [ ] Evidence class badges: substrate-capability vs interface-contract
|
||||||
|
- [ ] "Proposer was wrong" scenarios visually highlighted
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## Keyboard map (global, shipped with Wave 1)
|
||||||
|
|
||||||
|
| Shortcut | Action |
|
||||||
|
|---|---|
|
||||||
|
| `Cmd+K` | Command palette |
|
||||||
|
| `Cmd+I` | Toggle inspector |
|
||||||
|
| `Cmd+1`..`Cmd+0` | Navigate to route 1-10 |
|
||||||
|
| `j/k` or Up/Down | Navigate lists |
|
||||||
|
| `Enter` | Open selected item |
|
||||||
|
| `Esc` | Close drawer/palette/inspector |
|
||||||
|
| `/` | Focus search input |
|
||||||
|
| `?` | Show keyboard shortcut overlay |
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## Dependencies
|
||||||
|
|
||||||
|
```
|
||||||
|
Wave 1 (evidence spine) --> Wave 2 (projections, parallelizable)
|
||||||
|
--> Wave 3 (polish + demos)
|
||||||
|
```
|
||||||
|
|
||||||
|
Wave 1 must complete before Wave 2 work begins. Wave 2 routes are independent
|
||||||
|
of each other. Wave 3 can overlap with late Wave 2 work.
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## Explicit exclusions
|
||||||
|
|
||||||
|
Per ADR-0160 doctrine and CLAUDE.md:
|
||||||
|
|
||||||
|
- No multi-user auth, cloud deployment, or SaaS surface
|
||||||
|
- No animated "thinking" indicators or decorative motion
|
||||||
|
- No dashboard analytics walls
|
||||||
|
- No plugin/agent marketplace
|
||||||
|
- No corpus/pack mutation from the UI
|
||||||
|
- No mobile layout (engineering workstation only)
|
||||||
|
- No Deephaven, heavy JVM dependencies, or streaming database engines
|
||||||
|
- No framework upgrades (stays React 18 + stdlib HTTP)
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## Peer review record
|
||||||
|
|
||||||
|
| Reviewer | Key contribution |
|
||||||
|
|---|---|
|
||||||
|
| Claude Opus 4.6 | Original 7-phase plan; synthesis into evidence spine after critique |
|
||||||
|
| Codex | Path selection: "evidence spine first" over routes-first; evidence chain as intrinsic UI manifold; mutation doctrine correction (admitted corridors, not "no buttons") |
|
||||||
|
| GPT5.5-Thinking | Trace honesty correction: workbench chat turns do not attach telemetry sink; `ChatTurnResult` content is not in runtime telemetry; solution = Workbench Turn Evidence Journal as honest read model |
|
||||||
201
tests/test_workbench_journal.py
Normal file
201
tests/test_workbench_journal.py
Normal file
|
|
@ -0,0 +1,201 @@
|
||||||
|
from __future__ import annotations
|
||||||
|
|
||||||
|
import json
|
||||||
|
from dataclasses import replace
|
||||||
|
from pathlib import Path
|
||||||
|
|
||||||
|
import pytest
|
||||||
|
|
||||||
|
from workbench import api as workbench_api
|
||||||
|
from workbench.api import MAX_CHAT_PROMPT_CHARS, WorkbenchApi
|
||||||
|
from workbench.journal import TurnJournal, TurnJournalEntry
|
||||||
|
from workbench.schemas import ChatTurnResult, TurnVerdict
|
||||||
|
|
||||||
|
|
||||||
|
def _chat_result(prompt: str = "What is truth?") -> ChatTurnResult:
|
||||||
|
return ChatTurnResult(
|
||||||
|
prompt=prompt,
|
||||||
|
surface="Truth is coherent structure.",
|
||||||
|
articulation_surface="Truth is coherent structure.",
|
||||||
|
walk_surface="truth -> coherence",
|
||||||
|
grounding_source="pack",
|
||||||
|
epistemic_state="decoded",
|
||||||
|
normative_clearance="cleared",
|
||||||
|
normative_detail="",
|
||||||
|
trace_hash="sha256:trace",
|
||||||
|
refusal_emitted=False,
|
||||||
|
hedge_injected=False,
|
||||||
|
mutation_mode="runtime_turn",
|
||||||
|
identity_verdict=TurnVerdict(outcome="cleared", runtime_detail=""),
|
||||||
|
safety_verdict=TurnVerdict(outcome="cleared", runtime_detail=""),
|
||||||
|
ethics_verdict=TurnVerdict(outcome="cleared", runtime_detail=""),
|
||||||
|
proposal_candidates=[],
|
||||||
|
turn_cost_ms=7,
|
||||||
|
checkpoint_emitted=False,
|
||||||
|
)
|
||||||
|
|
||||||
|
|
||||||
|
def _entry(turn_id: int = 1, prompt: str = "What is truth?") -> TurnJournalEntry:
|
||||||
|
result = replace(_chat_result(prompt), turn_id=turn_id)
|
||||||
|
return TurnJournalEntry.from_chat_turn(
|
||||||
|
result,
|
||||||
|
turn_id=turn_id,
|
||||||
|
timestamp="2026-06-12T00:00:00+00:00",
|
||||||
|
)
|
||||||
|
|
||||||
|
|
||||||
|
def _request(api: WorkbenchApi, method: str, path: str, body: dict | None = None):
|
||||||
|
raw = b"" if body is None else json.dumps(body).encode("utf-8")
|
||||||
|
return api.handle(method, path, raw)
|
||||||
|
|
||||||
|
|
||||||
|
def _snapshot(root: Path) -> dict[str, bytes]:
|
||||||
|
snap: dict[str, bytes] = {}
|
||||||
|
if not root.exists():
|
||||||
|
return snap
|
||||||
|
for path in sorted(root.rglob("*")):
|
||||||
|
if path.is_file() and "__pycache__" not in path.relative_to(root).parts:
|
||||||
|
snap[path.relative_to(root).as_posix()] = path.read_bytes()
|
||||||
|
return snap
|
||||||
|
|
||||||
|
|
||||||
|
def test_journal_appends_without_modifying_existing_entries(tmp_path: Path) -> None:
|
||||||
|
journal = TurnJournal(tmp_path / "workbench_data")
|
||||||
|
|
||||||
|
first = journal.append(_entry(1))
|
||||||
|
first_line = journal.path.read_text(encoding="utf-8").splitlines()[0]
|
||||||
|
second = journal.append(_entry(2, "What is memory?"))
|
||||||
|
|
||||||
|
lines = journal.path.read_text(encoding="utf-8").splitlines()
|
||||||
|
assert len(lines) == 2
|
||||||
|
assert lines[0] == first_line
|
||||||
|
assert first.journal_digest
|
||||||
|
assert second.turn_id == 2
|
||||||
|
|
||||||
|
|
||||||
|
def test_journal_ordering_and_pagination_are_sequential(tmp_path: Path) -> None:
|
||||||
|
journal = TurnJournal(tmp_path / "workbench_data")
|
||||||
|
for turn_id in range(1, 5):
|
||||||
|
journal.append(_entry(turn_id, f"prompt {turn_id}"))
|
||||||
|
|
||||||
|
assert journal.next_turn_id() == 5
|
||||||
|
page = journal.list_summaries(limit=2, offset=1)
|
||||||
|
assert [item.turn_id for item in page] == [2, 3]
|
||||||
|
|
||||||
|
|
||||||
|
def test_trace_turns_offset_beyond_journal_length_returns_empty_items(
|
||||||
|
tmp_path: Path,
|
||||||
|
) -> None:
|
||||||
|
journal = TurnJournal(tmp_path / "workbench_data")
|
||||||
|
journal.append(_entry(1))
|
||||||
|
api = WorkbenchApi(journal=journal)
|
||||||
|
|
||||||
|
response = _request(api, "GET", "/trace/turns?limit=50&offset=50")
|
||||||
|
|
||||||
|
assert response.status == 200
|
||||||
|
assert response.payload["data"]["items"] == []
|
||||||
|
|
||||||
|
|
||||||
|
def test_empty_journal_returns_empty_items(tmp_path: Path) -> None:
|
||||||
|
api = WorkbenchApi(journal_dir=tmp_path / "workbench_data")
|
||||||
|
|
||||||
|
response = _request(api, "GET", "/trace/turns")
|
||||||
|
|
||||||
|
assert response.status == 200
|
||||||
|
assert response.payload["data"]["items"] == []
|
||||||
|
|
||||||
|
|
||||||
|
def test_prompt_size_limit_is_enforced_before_journaling(tmp_path: Path) -> None:
|
||||||
|
api = WorkbenchApi(journal_dir=tmp_path / "workbench_data")
|
||||||
|
prompt = "x" * (MAX_CHAT_PROMPT_CHARS + 1)
|
||||||
|
|
||||||
|
response = _request(api, "POST", "/chat/turn", {"prompt": prompt})
|
||||||
|
|
||||||
|
assert response.status == 400
|
||||||
|
assert not (tmp_path / "workbench_data" / "turn_journal.jsonl").exists()
|
||||||
|
|
||||||
|
|
||||||
|
def test_journal_rejects_paths_outside_workbench_data(tmp_path: Path) -> None:
|
||||||
|
with pytest.raises(ValueError, match="workbench_data"):
|
||||||
|
TurnJournal(tmp_path / "not_workbench_data")
|
||||||
|
|
||||||
|
|
||||||
|
def test_journal_does_not_write_teaching_pack_or_engine_state_roots(
|
||||||
|
tmp_path: Path,
|
||||||
|
monkeypatch: pytest.MonkeyPatch,
|
||||||
|
) -> None:
|
||||||
|
repo_root = Path(__file__).resolve().parent.parent
|
||||||
|
guarded = {
|
||||||
|
"teaching": repo_root / "teaching",
|
||||||
|
"packs": repo_root / "packs",
|
||||||
|
"language_packs/data": repo_root / "language_packs" / "data",
|
||||||
|
"engine_state": repo_root / "engine_state",
|
||||||
|
}
|
||||||
|
before = {name: _snapshot(path) for name, path in guarded.items()}
|
||||||
|
|
||||||
|
def fake_run(prompt: str) -> ChatTurnResult:
|
||||||
|
return _chat_result(prompt)
|
||||||
|
|
||||||
|
monkeypatch.setattr(workbench_api, "_run_chat_turn", fake_run)
|
||||||
|
api = WorkbenchApi(journal_dir=tmp_path / "workbench_data")
|
||||||
|
|
||||||
|
response = _request(api, "POST", "/chat/turn", {"prompt": "What is truth?"})
|
||||||
|
|
||||||
|
assert response.status == 200
|
||||||
|
assert {name: _snapshot(path) for name, path in guarded.items()} == before
|
||||||
|
assert api._journal.path.exists() # noqa: SLF001 - verifies the configured boundary.
|
||||||
|
|
||||||
|
|
||||||
|
def test_journal_digest_is_deterministic_for_identical_content(tmp_path: Path) -> None:
|
||||||
|
journal = TurnJournal(tmp_path / "workbench_data")
|
||||||
|
first = journal.append(_entry(1))
|
||||||
|
|
||||||
|
other_journal = TurnJournal(tmp_path / "other" / "workbench_data")
|
||||||
|
second = other_journal.append(_entry(1))
|
||||||
|
|
||||||
|
assert first.journal_digest == second.journal_digest
|
||||||
|
|
||||||
|
|
||||||
|
def test_chat_turn_round_trips_through_trace_endpoint(
|
||||||
|
tmp_path: Path,
|
||||||
|
monkeypatch: pytest.MonkeyPatch,
|
||||||
|
) -> None:
|
||||||
|
def fake_run(prompt: str) -> ChatTurnResult:
|
||||||
|
return _chat_result(prompt)
|
||||||
|
|
||||||
|
monkeypatch.setattr(workbench_api, "_run_chat_turn", fake_run)
|
||||||
|
api = WorkbenchApi(journal_dir=tmp_path / "workbench_data")
|
||||||
|
|
||||||
|
chat = _request(api, "POST", "/chat/turn", {"prompt": "What is truth?"})
|
||||||
|
turn_id = chat.payload["data"]["turn_id"]
|
||||||
|
trace = _request(api, "GET", f"/trace/{turn_id}")
|
||||||
|
|
||||||
|
assert chat.status == 200
|
||||||
|
assert trace.status == 200
|
||||||
|
for field in [
|
||||||
|
"turn_id",
|
||||||
|
"prompt",
|
||||||
|
"surface",
|
||||||
|
"articulation_surface",
|
||||||
|
"walk_surface",
|
||||||
|
"trace_hash",
|
||||||
|
"grounding_source",
|
||||||
|
"epistemic_state",
|
||||||
|
"normative_clearance",
|
||||||
|
"refusal_emitted",
|
||||||
|
"hedge_injected",
|
||||||
|
"proposal_candidates",
|
||||||
|
"turn_cost_ms",
|
||||||
|
"checkpoint_emitted",
|
||||||
|
]:
|
||||||
|
assert trace.payload["data"][field] == chat.payload["data"][field]
|
||||||
|
|
||||||
|
|
||||||
|
def test_unknown_turn_returns_404(tmp_path: Path) -> None:
|
||||||
|
api = WorkbenchApi(journal_dir=tmp_path / "workbench_data")
|
||||||
|
|
||||||
|
missing = _request(api, "GET", "/trace/999")
|
||||||
|
invalid = _request(api, "GET", "/trace/not-a-real-turn")
|
||||||
|
|
||||||
|
assert missing.status == 404
|
||||||
|
assert invalid.status == 404
|
||||||
|
|
@ -6,6 +6,7 @@ import json
|
||||||
import threading
|
import threading
|
||||||
import time
|
import time
|
||||||
from dataclasses import dataclass
|
from dataclasses import dataclass
|
||||||
|
from pathlib import Path
|
||||||
from typing import Any
|
from typing import Any
|
||||||
from urllib.parse import parse_qs, unquote, urlparse
|
from urllib.parse import parse_qs, unquote, urlparse
|
||||||
|
|
||||||
|
|
@ -17,6 +18,7 @@ from core.epistemic_state import (
|
||||||
normative_detail_from_verdicts,
|
normative_detail_from_verdicts,
|
||||||
)
|
)
|
||||||
from workbench import readers
|
from workbench import readers
|
||||||
|
from workbench.journal import DEFAULT_JOURNAL_DIR, TurnJournal, TurnJournalEntry
|
||||||
from workbench.readers import ArtifactTooLargeError
|
from workbench.readers import ArtifactTooLargeError
|
||||||
from workbench.schemas import ChatTurnResult, MathRatifyResult, ProposalRef, TurnVerdict, error, ok
|
from workbench.schemas import ChatTurnResult, MathRatifyResult, ProposalRef, TurnVerdict, error, ok
|
||||||
|
|
||||||
|
|
@ -33,8 +35,17 @@ class ApiResponse:
|
||||||
|
|
||||||
|
|
||||||
class WorkbenchApi:
|
class WorkbenchApi:
|
||||||
def __init__(self, telemetry_sink: Any | None = None) -> None:
|
def __init__(
|
||||||
|
self,
|
||||||
|
telemetry_sink: Any | None = None,
|
||||||
|
*,
|
||||||
|
journal: TurnJournal | None = None,
|
||||||
|
journal_dir: Any | None = None,
|
||||||
|
) -> None:
|
||||||
self._telemetry_sink = telemetry_sink
|
self._telemetry_sink = telemetry_sink
|
||||||
|
self._journal = journal or TurnJournal(
|
||||||
|
DEFAULT_JOURNAL_DIR if journal_dir is None else Path(journal_dir)
|
||||||
|
)
|
||||||
|
|
||||||
def attach_telemetry_sink(self, sink: Any | None) -> None:
|
def attach_telemetry_sink(self, sink: Any | None) -> None:
|
||||||
self._telemetry_sink = sink
|
self._telemetry_sink = sink
|
||||||
|
|
@ -145,8 +156,21 @@ class WorkbenchApi:
|
||||||
return ApiResponse(200, ok(result))
|
return ApiResponse(200, ok(result))
|
||||||
if method == "POST" and path == "/chat/turn":
|
if method == "POST" and path == "/chat/turn":
|
||||||
return self._chat_turn(body)
|
return self._chat_turn(body)
|
||||||
|
if method == "GET" and path == "/trace/turns":
|
||||||
|
limit = int(query.get("limit", ["50"])[0])
|
||||||
|
offset = int(query.get("offset", ["0"])[0])
|
||||||
|
items = self._journal.list_summaries(limit=limit, offset=offset)
|
||||||
|
return ApiResponse(200, ok({"items": items}))
|
||||||
if method == "GET" and path.startswith("/trace/"):
|
if method == "GET" and path.startswith("/trace/"):
|
||||||
return ApiResponse(404, error("not_found", "trace storage is not wired in W-026"))
|
raw_turn_id = unquote(path.removeprefix("/trace/"))
|
||||||
|
try:
|
||||||
|
turn_id = int(raw_turn_id)
|
||||||
|
except ValueError:
|
||||||
|
return ApiResponse(404, error("not_found", f"trace turn not found: {raw_turn_id}"))
|
||||||
|
try:
|
||||||
|
return ApiResponse(200, ok(self._journal.get_entry(turn_id)))
|
||||||
|
except FileNotFoundError:
|
||||||
|
return ApiResponse(404, error("not_found", f"trace turn not found: {turn_id}"))
|
||||||
if method == "GET" and path.startswith("/replay/"):
|
if method == "GET" and path.startswith("/replay/"):
|
||||||
return ApiResponse(501, error("unsupported", "route is deferred beyond W-026"))
|
return ApiResponse(501, error("unsupported", "route is deferred beyond W-026"))
|
||||||
return ApiResponse(404, error("not_found", f"route not found: {method} {path}"))
|
return ApiResponse(404, error("not_found", f"route not found: {method} {path}"))
|
||||||
|
|
@ -278,13 +302,21 @@ class WorkbenchApi:
|
||||||
started = time.perf_counter()
|
started = time.perf_counter()
|
||||||
result = _run_chat_turn(prompt)
|
result = _run_chat_turn(prompt)
|
||||||
elapsed_ms = max(0, int(round((time.perf_counter() - started) * 1000)))
|
elapsed_ms = max(0, int(round((time.perf_counter() - started) * 1000)))
|
||||||
return ApiResponse(200, ok(_with_turn_cost(result, elapsed_ms)))
|
turn_id = self._journal.next_turn_id()
|
||||||
|
result_with_cost = _with_turn_cost_and_id(result, elapsed_ms, turn_id)
|
||||||
|
entry = TurnJournalEntry.from_chat_turn(result_with_cost, turn_id=turn_id)
|
||||||
|
self._journal.append(entry)
|
||||||
|
return ApiResponse(200, ok(result_with_cost))
|
||||||
|
|
||||||
|
|
||||||
def _with_turn_cost(result: ChatTurnResult, turn_cost_ms: int) -> ChatTurnResult:
|
def _with_turn_cost_and_id(
|
||||||
|
result: ChatTurnResult,
|
||||||
|
turn_cost_ms: int,
|
||||||
|
turn_id: int,
|
||||||
|
) -> ChatTurnResult:
|
||||||
from dataclasses import replace
|
from dataclasses import replace
|
||||||
|
|
||||||
return replace(result, turn_cost_ms=turn_cost_ms)
|
return replace(result, turn_cost_ms=turn_cost_ms, turn_id=turn_id)
|
||||||
|
|
||||||
|
|
||||||
def _coerce_grounding_source(value: object) -> str:
|
def _coerce_grounding_source(value: object) -> str:
|
||||||
|
|
|
||||||
174
workbench/journal.py
Normal file
174
workbench/journal.py
Normal file
|
|
@ -0,0 +1,174 @@
|
||||||
|
"""Append-only Workbench turn evidence journal."""
|
||||||
|
|
||||||
|
from __future__ import annotations
|
||||||
|
|
||||||
|
import hashlib
|
||||||
|
import json
|
||||||
|
import threading
|
||||||
|
from dataclasses import dataclass, replace
|
||||||
|
from pathlib import Path
|
||||||
|
from typing import Any
|
||||||
|
|
||||||
|
from workbench.schemas import ChatTurnResult, to_data, utc_now
|
||||||
|
|
||||||
|
|
||||||
|
REPO_ROOT = Path(__file__).resolve().parents[1]
|
||||||
|
DEFAULT_JOURNAL_DIR = REPO_ROOT / "workbench_data"
|
||||||
|
JOURNAL_FILENAME = "turn_journal.jsonl"
|
||||||
|
PROMPT_EXCERPT_CHARS = 120
|
||||||
|
SURFACE_EXCERPT_CHARS = 120
|
||||||
|
|
||||||
|
|
||||||
|
@dataclass(frozen=True, slots=True)
|
||||||
|
class TurnJournalSummary:
|
||||||
|
turn_id: int
|
||||||
|
timestamp: str
|
||||||
|
prompt_excerpt: str
|
||||||
|
surface_excerpt: str
|
||||||
|
trace_hash: str | None
|
||||||
|
grounding_source: str
|
||||||
|
|
||||||
|
|
||||||
|
@dataclass(frozen=True, slots=True)
|
||||||
|
class TurnJournalEntry:
|
||||||
|
turn_id: int
|
||||||
|
timestamp: str
|
||||||
|
trace_hash: str | None
|
||||||
|
prompt: str
|
||||||
|
surface: str
|
||||||
|
articulation_surface: str | None
|
||||||
|
walk_surface: str | None
|
||||||
|
grounding_source: str
|
||||||
|
epistemic_state: str
|
||||||
|
normative_clearance: str
|
||||||
|
verdicts: dict[str, Any]
|
||||||
|
refusal_emitted: bool
|
||||||
|
hedge_injected: bool
|
||||||
|
proposal_candidates: list[dict[str, Any]]
|
||||||
|
turn_cost_ms: int
|
||||||
|
checkpoint_emitted: bool
|
||||||
|
journal_digest: str = ""
|
||||||
|
|
||||||
|
@classmethod
|
||||||
|
def from_chat_turn(
|
||||||
|
cls,
|
||||||
|
result: ChatTurnResult,
|
||||||
|
*,
|
||||||
|
turn_id: int,
|
||||||
|
timestamp: str | None = None,
|
||||||
|
) -> "TurnJournalEntry":
|
||||||
|
return cls(
|
||||||
|
turn_id=turn_id,
|
||||||
|
timestamp=timestamp or utc_now(),
|
||||||
|
trace_hash=result.trace_hash,
|
||||||
|
prompt=result.prompt,
|
||||||
|
surface=result.surface,
|
||||||
|
articulation_surface=result.articulation_surface,
|
||||||
|
walk_surface=result.walk_surface,
|
||||||
|
grounding_source=result.grounding_source,
|
||||||
|
epistemic_state=result.epistemic_state,
|
||||||
|
normative_clearance=result.normative_clearance,
|
||||||
|
verdicts={
|
||||||
|
"identity": to_data(result.identity_verdict),
|
||||||
|
"safety": to_data(result.safety_verdict),
|
||||||
|
"ethics": to_data(result.ethics_verdict),
|
||||||
|
},
|
||||||
|
refusal_emitted=result.refusal_emitted,
|
||||||
|
hedge_injected=result.hedge_injected,
|
||||||
|
proposal_candidates=[
|
||||||
|
candidate for candidate in to_data(result.proposal_candidates)
|
||||||
|
],
|
||||||
|
turn_cost_ms=result.turn_cost_ms,
|
||||||
|
checkpoint_emitted=result.checkpoint_emitted,
|
||||||
|
)
|
||||||
|
|
||||||
|
def summary(self) -> TurnJournalSummary:
|
||||||
|
return TurnJournalSummary(
|
||||||
|
turn_id=self.turn_id,
|
||||||
|
timestamp=self.timestamp,
|
||||||
|
prompt_excerpt=self.prompt[:PROMPT_EXCERPT_CHARS],
|
||||||
|
surface_excerpt=self.surface[:SURFACE_EXCERPT_CHARS],
|
||||||
|
trace_hash=self.trace_hash,
|
||||||
|
grounding_source=self.grounding_source,
|
||||||
|
)
|
||||||
|
|
||||||
|
|
||||||
|
class TurnJournal:
|
||||||
|
"""Pure JSONL append/read model for Workbench chat evidence."""
|
||||||
|
|
||||||
|
def __init__(self, journal_dir: Path = DEFAULT_JOURNAL_DIR) -> None:
|
||||||
|
self._journal_dir = _validate_journal_dir(journal_dir)
|
||||||
|
self._path = self._journal_dir / JOURNAL_FILENAME
|
||||||
|
self._lock = threading.Lock()
|
||||||
|
|
||||||
|
@property
|
||||||
|
def journal_dir(self) -> Path:
|
||||||
|
return self._journal_dir
|
||||||
|
|
||||||
|
@property
|
||||||
|
def path(self) -> Path:
|
||||||
|
return self._path
|
||||||
|
|
||||||
|
def next_turn_id(self) -> int:
|
||||||
|
entries = self._read_entries()
|
||||||
|
if not entries:
|
||||||
|
return 1
|
||||||
|
return max(entry.turn_id for entry in entries) + 1
|
||||||
|
|
||||||
|
def append(self, entry: TurnJournalEntry) -> TurnJournalEntry:
|
||||||
|
with self._lock:
|
||||||
|
expected = self.next_turn_id()
|
||||||
|
if entry.turn_id != expected:
|
||||||
|
raise ValueError(
|
||||||
|
f"turn_id must be next sequential id {expected}, got {entry.turn_id}"
|
||||||
|
)
|
||||||
|
sealed = replace(entry, journal_digest=_journal_digest(entry))
|
||||||
|
self._journal_dir.mkdir(parents=True, exist_ok=True)
|
||||||
|
with self._path.open("a", encoding="utf-8") as fh:
|
||||||
|
fh.write(_canonical_json(to_data(sealed)))
|
||||||
|
fh.write("\n")
|
||||||
|
return sealed
|
||||||
|
|
||||||
|
def list_summaries(self, *, limit: int = 50, offset: int = 0) -> list[TurnJournalSummary]:
|
||||||
|
if limit < 0:
|
||||||
|
raise ValueError("limit must be non-negative")
|
||||||
|
if offset < 0:
|
||||||
|
raise ValueError("offset must be non-negative")
|
||||||
|
entries = self._read_entries()
|
||||||
|
return [entry.summary() for entry in entries[offset : offset + limit]]
|
||||||
|
|
||||||
|
def get_entry(self, turn_id: int) -> TurnJournalEntry:
|
||||||
|
for entry in self._read_entries():
|
||||||
|
if entry.turn_id == turn_id:
|
||||||
|
return entry
|
||||||
|
raise FileNotFoundError(str(turn_id))
|
||||||
|
|
||||||
|
def _read_entries(self) -> list[TurnJournalEntry]:
|
||||||
|
if not self._path.exists():
|
||||||
|
return []
|
||||||
|
entries: list[TurnJournalEntry] = []
|
||||||
|
with self._path.open("r", encoding="utf-8") as fh:
|
||||||
|
for line in fh:
|
||||||
|
if not line.strip():
|
||||||
|
continue
|
||||||
|
payload = json.loads(line)
|
||||||
|
entries.append(TurnJournalEntry(**payload))
|
||||||
|
return entries
|
||||||
|
|
||||||
|
|
||||||
|
def _validate_journal_dir(journal_dir: Path) -> Path:
|
||||||
|
resolved = journal_dir.resolve()
|
||||||
|
if resolved.name != "workbench_data":
|
||||||
|
raise ValueError("journal directory must be named workbench_data")
|
||||||
|
return resolved
|
||||||
|
|
||||||
|
|
||||||
|
def _canonical_json(payload: dict[str, Any]) -> str:
|
||||||
|
return json.dumps(payload, ensure_ascii=False, sort_keys=True, separators=(",", ":"))
|
||||||
|
|
||||||
|
|
||||||
|
def _journal_digest(entry: TurnJournalEntry) -> str:
|
||||||
|
payload = to_data(replace(entry, journal_digest=""))
|
||||||
|
payload.pop("journal_digest", None)
|
||||||
|
raw = _canonical_json(payload).encode("utf-8")
|
||||||
|
return "sha256:" + hashlib.sha256(raw).hexdigest()
|
||||||
|
|
@ -113,6 +113,38 @@ class ChatTurnResult:
|
||||||
proposal_candidates: list[ProposalRef]
|
proposal_candidates: list[ProposalRef]
|
||||||
turn_cost_ms: int
|
turn_cost_ms: int
|
||||||
checkpoint_emitted: bool
|
checkpoint_emitted: bool
|
||||||
|
turn_id: int | None = None
|
||||||
|
|
||||||
|
|
||||||
|
@dataclass(frozen=True, slots=True)
|
||||||
|
class TurnJournalSummarySchema:
|
||||||
|
turn_id: int
|
||||||
|
timestamp: str
|
||||||
|
prompt_excerpt: str
|
||||||
|
surface_excerpt: str
|
||||||
|
trace_hash: str | None
|
||||||
|
grounding_source: GroundingSource
|
||||||
|
|
||||||
|
|
||||||
|
@dataclass(frozen=True, slots=True)
|
||||||
|
class TurnJournalEntrySchema:
|
||||||
|
turn_id: int
|
||||||
|
timestamp: str
|
||||||
|
trace_hash: str | None
|
||||||
|
prompt: str
|
||||||
|
surface: str
|
||||||
|
articulation_surface: str | None
|
||||||
|
walk_surface: str | None
|
||||||
|
grounding_source: GroundingSource
|
||||||
|
epistemic_state: EpistemicStateValue
|
||||||
|
normative_clearance: NormativeClearanceValue
|
||||||
|
verdicts: dict[str, Any]
|
||||||
|
refusal_emitted: bool
|
||||||
|
hedge_injected: bool
|
||||||
|
proposal_candidates: list[dict[str, Any]]
|
||||||
|
turn_cost_ms: int
|
||||||
|
checkpoint_emitted: bool
|
||||||
|
journal_digest: str
|
||||||
|
|
||||||
|
|
||||||
@dataclass(frozen=True, slots=True)
|
@dataclass(frozen=True, slots=True)
|
||||||
|
|
|
||||||
9
workbench_data/README.md
Normal file
9
workbench_data/README.md
Normal file
|
|
@ -0,0 +1,9 @@
|
||||||
|
# Workbench Data
|
||||||
|
|
||||||
|
`turn_journal.jsonl` is a local, content-bearing evidence journal for CORE
|
||||||
|
Workbench chat turns. Entries contain user prompts and engine surfaces already
|
||||||
|
returned to the local operator by `/chat/turn`.
|
||||||
|
|
||||||
|
This directory is a Workbench read model only. It is not teaching memory,
|
||||||
|
runtime memory, semantic pack data, or a cognitive runtime fork. Generated JSONL
|
||||||
|
files are intentionally ignored by git.
|
||||||
Loading…
Reference in a new issue