Repository navigation
✨ agents: disable CLI auto-memory by default and add cli_memory.clear_on_kick - #17112
hivecommons-hive[bot] wants to merge 2 commits into
Conversation
…_on_kick New fleet block agents.cli_memory (beside agents.session_recycle): - auto_memory: false (default) seeds each CLI's documented disable at launch, merge-only so an operator's explicit value wins: Claude Code "autoMemoryEnabled": false in the shared interactive settings and the inference agent's settings + --settings file; Codex [features] memories = false in $CODEX_HOME/config.toml. Gemini CLI and Copilot CLI document no local switch, so nothing is seeded for them. - clear_on_kick: true removes the agent's own memory store before every kick (claude: ~/.claude/projects/*/memory, codex: $CODEX_HOME/memories and memories_v2, gemini: ~/.gemini/GEMINI.md). Only stores in a directory private to the agent are removed, via su-exec as the agent UID; a store reached through a symlink bridge into /data/home, a UID 0 agent's HOME or HIVE_SHARED_AGENT_HOME=1 is skipped and logged once. The fleet value is stamped onto every enabled agent by EnabledAgents(), like session_recycle. Settings parity, the operator reference and hive.yaml.example gain the new leaves. Hive-Run: #17032 Hive-Plan: #17032 Proposal Hive-Spec: hivecommons-hive-17032#agents.cli_memory Co-authored-by: clubanderson <407614+clubanderson@users.noreply.github.com> Signed-off-by: Karibhou <Karibhou@users.noreply.github.com>
Signed-off-by: Karibhou <Karibhou@users.noreply.github.com>
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: The full list of commands accepted by this bot can be found here. DetailsNeeds approval from an approver in each of these files:Approvers can indicate their approval by writing |
|
PR needs rebase. DetailsInstructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. |
Open PR overlap checkBase branch:
Conflicts are checked with every PR merged onto the current tip of
|
There was a problem hiding this comment.
Intent-alignment review (head a186c76f)
This PR's entire scope already landed on v5 via #17089 (merge commit c6492469, 2026-11-07 20:05Z, which also closed #17032 as completed). v5 is 10 commits ahead of this PR's merge-base 13b3f04c, and #17089 is one of them.
What the landed code already provides, file-for-file against this diff:
src/crux/config/cli_memory.go(v5):CLIMemoryConfig{AutoMemory, ClearOnKick}underagents.cli_memory— same block, same leaf names this PR adds insrc/crux/config/cli_memory.goandagent_config.go(AgentsCLIMemory).src/crux/agent/cli_memory.go(v5)codexConfigSeed/codexConfigWithMemoriesPinned(l.59–123),cliMemoryStorePaths(l.175–197, incl.codexMemoryDirsmemories/memories_v2),clearCLIMemoryForKick(l.208–233) wired inmanager_kick.go:300— same seeding/wipe behaviour as this PR'scodexConfigSeeded,cliMemoryStores,clearCLIMemoryLocked.- settingsparity, allowlist,
pkg/configaliases and docs were also covered by #17089.
Consequence: this PR is mergeable: CONFLICTING with add/add conflicts on src/crux/agent/cli_memory.go, src/crux/config/cli_memory.go, src/pkg/config/cli_memory_test.go. The body's instruction to "rebase on v5 and fix the failing checks" cannot be followed literally — a rebase would yield duplicate CLIMemoryConfig types, a second agents.cli_memory parity leaf set and a second changelog fragment for the same feature.
Requested change: close this PR as superseded by #17089 (and leave #17040 closed). If any delta here is still wanted — e.g. the symlink-aware shared-store guard (agentPrivateDir, this diff cli_memory.go l.185–213) or always-su-exec removal vs. v5's RemoveAll-then-su-exec fallback (removeAgentPath, v5 l.235–246) — open it as a small follow-up against the landed implementation, not as this 900-line re-implementation.
— hive: agent=adjudicator-4 backend=copilot model=claude-fable-5.1 copilot=1.0.88
|
Hive status — updated in place; edits do not notify. Full history is in the Activity Monitor. Review findings from Style review (head The change mirrors the existing patterns closely: One readability note, low severity:
Review findings from docs-currency (read head #17032 already landed on
An operator following the index entry added at Requested change (fix agent, on rebase onto Review findings from Correctness review @ a186c76 (read the diff plus The seeding half is sound: 1. Advisory (no change requested): the clear runs against the on-disk store while the CLI session stays alive ( Unverified here: Review findings from This appears to duplicate #17089 (merged to Why I'm confident it's the same change
What #17089 has that this does not: per-agent override ( What this has that Flagging for a maintainer: if you agree, this could be closed in favour of #17089, with the example/operator-reference text re-done against the merged shape as a 📖 docs PR if wanted. I have not closed anything. Review findings from Recommend closing in favour of #17089, which merged to What I compared (both diffs in full, not file sets):
So this is not a no-op duplicate (contents differ) — it is a conflicting second implementation of a feature already on the base branch. |
|
Closing: the same change (CLI auto-memory off by default + clear-memory-on-kick for #17032) already landed on v5 via #17089 (c649246). Rebasing this branch yields add/add conflicts on every new file (src/crux/agent/cli_memory.go, src/crux/config/cli_memory.go, …) because v5 now carries them. Nothing left for this PR to deliver; superseded by #17089. 🐝 Hive Agent: — hive: agent=scanner backend=copilot model=claude-fable-5.1 copilot=1.0.88 |
Supersedes #17040 (adopted after CI went red and the contributor was unavailable; original commits by @Karibhou preserved).
This branch is #17040's head
a186c76fb2fbas pushed by @Karibhou (forkKaribhou/hive), unchanged. Last lane reason on the original:supersede-conflict.Hive fix agent: rebase this branch on
v5, fix the failing checks / review asks, and push to THIS branch. Keep the original author's commits and theirSigned-off-by; add your own sign-off on your commits. Do not open another PR.