This release adds 3 notable features for engineering teams evaluating rollout.
✓ No known CVEs patched in this version
Topics
Summary
AI summaryFixed hangs on stdio transport and oversized response handling.
Changes in this release
| Type | Severity | Summary | CVE |
|---|---|---|---|
| Feature | Low |
Uses stall watchdogs instead of total‑time caps for seed/clone operations. Uses stall watchdogs instead of total‑time caps for seed/clone operations. Source: llm_adapter@2026-06-10 Confidence: high |
— |
| Feature | Low |
Aligns initial‑scan log markers to `indexing`/`indexed` states for parity. Aligns initial‑scan log markers to `indexing`/`indexed` states for parity. Source: llm_adapter@2026-06-10 Confidence: high |
— |
| Feature | Low |
Moves bulk passthrough fields to temporary files referenced by `reportPath` to stay within token budgets. Moves bulk passthrough fields to temporary files referenced by `reportPath` to stay within token budgets. Source: llm_adapter@2026-06-10 Confidence: high |
— |
| Bugfix | Medium |
Prevents subprocess hang by setting stdin=DEVNULL unless input is piped. Prevents subprocess hang by setting stdin=DEVNULL unless input is piped. Source: llm_adapter@2026-06-10 Confidence: high |
— |
| Bugfix | Medium |
Ensures runner image tag updates correctly during upgrades. Ensures runner image tag updates correctly during upgrades. Source: llm_adapter@2026-06-10 Confidence: high |
— |
| Bugfix | Medium |
Prevents crash‑looping watchers from being counted as ready. Prevents crash‑looping watchers from being counted as ready. Source: llm_adapter@2026-06-10 Confidence: high |
— |
| Other | Low |
schema_version":1,"prompt_version":1}]<|fim_middle|><|fim_suffix|>eports/<tool>/` files referenced by an inline `reportPath`. Inline payloads stay compact under budget tests; reports are pruned at write time (keep last 5, max 7 days) and password-redacted. schema_version":1,"prompt_version":1}]<|fim_middle|><|fim_suffix|>eports/<tool>/` files referenced by an inline `reportPath`. Inline payloads stay compact under budget tests; reports are pruned at write time (keep last 5, max 7 days) and password-redacted. Source: llm_adapter@2026-06-10 Confidence: low |
— |
| Other | Low |
affected_surface affected_surface Source: llm_adapter@2026-06-10 Confidence: low |
— |
Full changelog
Patch release on the 2.5.0 readiness work, fixing the two remaining classes of tool misbehavior: hangs and oversized responses.
Fixed
- Carryover/tool hang on stdio transport (#49): subprocess children inherited the MCP server's stdin — which on stdio transport is the JSON-RPC pipe — wedging tool calls while the work completed server-side. Subprocesses now get
stdin=DEVNULLunless input is piped, proven by a before/after stdio regression harness (120s hang → 3.4s). A package-wide AST hygiene test prevents reintroduction. - Runner image tag drift on upgrade (#50):
providers/settings.pyderived the CGC runner image independently and dropped the layer revision, so upgrading hosts kept a cached guard-less image. The tag now comes from the singlecgc_runner_image()derivation, with a regression test. - Crash-looping watchers reported ready:
restartingcontainers no longer count toward readiness.
Changed
- Stall watchdogs instead of total-time caps for seed/clone: GrepAI index clones are killed only after 300s of zero progress (
GREPAI_CLONE_STALL_SECONDS), never for size — copying index data instead of re-indexing is what makes rapid worktree provider deployment viable, and it scales with index size by design. - GrepAI indexing-state parity: initial-scan log markers now map to
indexing/indexedinstead ofunknown, feeding the context-packetindexingbusy list. - Response token budgets: bulk passthrough fields (raw status trees, rebind transcripts, watcher step results) move to
temp/tool-reports/<tool>/files referenced by an inlinereportPath. Inline payloads stay compact under budget tests; reports are pruned at write time (keep last 5, max 7 days) and password-redacted.
Full changes: https://github.com/Foxfire1st/agents-remember-md/pull/51
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
Share this release
About Agents Remember
All releases →Related context
Related tools
Earlier breaking changes
- vmcp-v2.9.0 Removed tools direct_closeout_preview and direct_closeout_apply.
- vmcp-v2.0.0 `WorkflowKind` enum drops `heavy` and `heavy-task` values.
- vmcp-v2.0.0 `skills_install` no longer accepts a `layout` input; installer performs flat copy.
- vmcp-v2.0.0 Removed workflows W-01-heavy-task-workflow and W-03-chat-task-workflow.
Beta — feedback welcome: [email protected]