Pine/deepseek-harness
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases 1 Wiki Activity

docs: purge chain-of-thought leakage from prose

Browse Source 
Delete design-session citations (decision/audit/plan ordinals, stack
positions), change narration, review choreography, and reviewer-addressed
justification from comments, JSDoc, docs, READMEs, Agent Notes, tests, and
generator templates; restate every affected fact as current-state contract
prose. Fix generated docs at their sources and regenerate the catalogs and
cordis-surface regions; re-paste type-equiv blocks; update every bilingual
counterpart and re-record the pairs. Record the citation rule in the
committed-artifact-citations Agent Note.
This commit is contained in:
Tianyi Cui
2026-08-09 15:09:19 +08:00
parent 793f6f55df
commit 25dcd7293c
763 changed files with 2705 additions and 1710 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.agents/notes/implemented/simplification/2026-06-20-collapse-trace-only-session-events.md
+2 -6
View File
@@ -15,7 +15,7 @@ These events make the canonical transcript look more useful as telemetry than it
Standalone trace-only events are removed exactly where their information is preserved without a parallel record:
- Successful-step usage folds into the matching `assistant/message` (`assistant/message { turn, step, content, usage? }`), so the assembled model output and its accounting travel together.
- A failed or aborted step that has usage but no assistant content carries the usage on an empty-content `assistant/message` (the implementation note below carries the no-information-loss proof) — no persisted usage chunk goes unrepresented.
- A failed or aborted step that has usage but no assistant content carries the usage on an empty-content `assistant/message { content: [], usage }` — no persisted usage chunk goes unrepresented. The no-information-loss case is the max-tokens path: a step cut off with usage but empty content (e.g. only a dropped tool call) previously emitted a standalone `usage`. To keep the empty-content event from injecting a spurious content-less assistant turn into the provider transcript, `deriveMessages()` skips empty-content `assistant/message` events; a regression test asserts usage stays represented AND derived history stays uncorrupted.
- The step number from the standalone `error` event folds into `turn/end.reason` for `kind: 'error'` (`{ kind: 'error', step, message, code? }`) — `turn/end` is the durable turn outcome ACP and resume already consume.
- `agent/error` and logging stay for live diagnostics; there is no second session-log error record after `turn/end`.
@@ -31,14 +31,10 @@ The user conversation log contains what is needed to render, resume, audit, and
## Consequences
A consumer can no longer filter the canonical log for standalone `usage` or step-level `error` rows. It must read those facts from the assistant/failure events that carry them. That is a reasonable simplification only if the implementing PR proves the same facts remain present; otherwise the standalone events should stay.
A consumer can no longer filter the canonical log for standalone `usage` or step-level `error` rows. It must read those facts from the assistant/failure events that carry them. That is a reasonable simplification because the same facts remain present, as the Verification section proves.
## Implementation note
Shipped as proposed, with one scope refinement (per AGENTS.md "Agent Notes are proposals, not golden truth"):
- **Empty-content `assistant/message` hosts usage with no data loss.** The proof the proposal demanded (no persisted usage chunk becomes unrepresented) lands on the max-tokens path: a step cut off with usage but empty content (e.g. only a dropped tool call) previously emitted a standalone `usage`. It now records an empty-content `assistant/message { content: [], usage }`. To keep that from injecting a spurious content-less assistant turn into the provider transcript, `deriveMessages()` skips empty-content `assistant/message` events. A regression test asserts usage stays represented AND derived history is uncorrupted.
**Format version.** This changes persisted events, but the pre-release session format remains pinned at `0` and rejects any other version without migration. `dsh-session` owns the constant used by writers and load validation. Monotonic format versions begin at the first release.
Usage is now observed on `assistant/message.usage`; an operational error's step on `turn/end.reason` for `kind: 'error'`. `agent/error` + logging are unchanged for live diagnostics.