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

test(examples): snapshot headless one-shot stream

Browse Source
This commit is contained in:
Tianyi Cui
2026-07-15 22:43:03 +08:00
parent 7278fc21d8
commit 5e3da065db
13 changed files with 251 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
examples/acp-agent/README.md
View File

@@ -33,7 +33,7 @@ The editor sets each session's `cwd` to the project it opens, and bash uses that
## Snapshot tests (record-once / replay-deterministic)
This example hosts the ACP snapshot suite. `dsh-llm-replay` reconstructs model streams from `assistant/chunk` events in each scenario's session JSONL, so replay is keyless. Recording runs the real agent and harvests that log; refresh keeps the committed transcript as mock input and rewrites current replay outputs. `replay.override.json` covers throw and hang cases that chunks cannot express, and an optional `workspace/` seeds files. The [snapshot RFC](../../docs/rfc/implemented/testing/2026-06-19-acp-snapshot-tests.md) owns the full design.
This example hosts the ACP snapshot suite and the headless `stream-json` snapshot. Both replay through `dsh-llm-replay`, which reconstructs model streams from `assistant/chunk` events in each scenario's session JSONL. The headless snapshot reuses `advanced-toolchain` to pin the one-shot stream plus its re-persisted parent and child logs; child activity appears in the stream only through parent tool events. Recording runs the real ACP agent and harvests its logs; refresh keeps the committed transcript as mock input and rewrites current replay outputs. `replay.override.json` covers throw and hang cases that chunks cannot express, and an optional `workspace/` seeds files. The [snapshot RFC](../../docs/rfc/implemented/testing/2026-06-19-acp-snapshot-tests.md) owns the ACP harness design.
## Permissions and sandboxing