Bump OpenHuman for prompt-cache and usage fixes

[senamakel]

Summary

Bumps vendor/openhuman from 4739e97284 to 1f2e59a238, the merge of tinyhumansai/openhuman#7443. That merge brings in two fixes found while measuring prompt caching on euler_hive runs:

• tinyinference Show a seat its own conversations when it answers inside one #82: Anthropic-family models now get a cache breakpoint on the conversation tail. In a tool loop, the last user message stops moving while assistant and tool-result messages build up after it. Without that tail breakpoint, every call re-sent the growth uncached. On a fixed 9-call tool loop through OpenRouter (anthropic/claude-haiku-5.5), cached prompt tokens went from 31% to 82%.
• tinyagents #387: BudgetMiddleware and the agent loop both emitted UsageRecorded for every model call, which doubled every journal token total. Hit ratios were unaffected.

The openhuman, tinytools and tinymemory rev pins in both manifests now match the vendored submodules.

No tinyhivemind code changes. euler_hive already appends only new desk rows per turn on a continuing session, so its prompt layout keeps a stable prefix.

Related issue

None.

API or behavior changes

None in this repository's crates. Behavior changes come from OpenHuman: Anthropic-family requests carry an extra cache breakpoint, and usage events are no longer duplicated.

Validation

• .github/scripts/assert-openhuman-pin.sh: all six pins match.
• cargo fmt --all -- --check: ok.
• [~] cargo clippy --all-targets --all-features -- -D warnings: clean for workspace crates. The only output is warnings from vendored OpenHuman sources, and the local-only rustc 1.96 doc_markdown lints in tinyhivemind-hives, which CI passes.
• cargo build --all-targets --all-features: ok.
• cargo test --all-features: all pass.
• .github/scripts/assert-pure.sh: clean.
• cargo clippy --manifest-path examples/openhuman/Cargo.toml --all-targets -- -D warnings: ok.
• cargo test --manifest-path examples/openhuman/Cargo.toml: all pass. The first run exited 101 while a live euler_hive run used the same machine, and the failing test name was not captured. A clean rerun passed every suite (51, 44, 16, 5, 2 and 1 tests).
• euler_hive built from this branch is running a live PE1012 run.

Checklist

• No new #[allow(...)] or #[ignore], and no relaxed lints.
• No secrets in the diff or the description.

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-11T15:00:58.661084Z	8ab6fd1	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

• Configuration used: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: 368663c6-7ffc-4132-b54e-2efaf31c43bc

---

---

📥 Commits

Reviewing files that changed from the base of the PR and between b822ac1 and 8ab6fd1.

---

---

⛔ Files ignored due to path filters (2)

• Cargo.lock is excluded by !**/*.lock
• examples/openhuman/Cargo.lock is excluded by !**/*.lock

---

---

📒 Files selected for processing (3)

• Cargo.toml
• examples/openhuman/Cargo.toml
• vendor/openhuman

---

---

Included review availability: This review used your included allowance. Your plan provides up to 1 included review per hour; 0 remain after this review.

---

---

---

📝 Walkthrough

📝 Walkthrough

Walkthrough

The Cargo manifests now pin OpenHuman and Tinytools to updated Git revisions. The vendor/openhuman subproject reference also points to the updated OpenHuman revision.

Changes

Dependency Revision Pins

Layer / File(s)	Summary
Update dependency revisions Cargo.toml, examples/openhuman/Cargo.toml, vendor/openhuman	The manifests pin OpenHuman and Tinytools to updated revisions. The vendor/openhuman subproject reference uses the updated OpenHuman revision.

Priority: ➖ Normal

Estimated code review effort: 1 (Trivial) | ~5 minutes

Change: Bug fix

Suggested reviewers: sanil-23

---

---

Merge Risk: ⚪ Minimal · up to 8ab6f

The updated dependency revisions are aligned across the manifests and vendored OpenHuman reference; no concrete merge-blocking risk is indicated.

Architecture Summary

Architecture risk: 🔵 Low · up to 8ab6f

The change affects 2 systems.

Changed systems: Cargo.toml, examples

Architecture concerns
No architecture-level concerns identified.

Review details

Systems and components

• observed — Cargo.toml (service) was modified; 1 changed file maps to changed impact.
• observed — examples (service) was modified; 1 changed file maps to changed impact.

Before / after behavior

• observed — Modified behavior in Cargo.toml: The OpenHuman dependencies change from revision 467ab703acccb63391a8dd6a530297ef904dda76 to 1f2e59a238c3221004aabda3e5facc2f5624eaa6; both Tinytools dependencies change from revision 360721bc3c7a26221b61f6629f5f0415cfccb568 to 1cd9e9c4a117336e6f42d0c7936ed60e7476d88f.
• observed — Modified behavior in examples/openhuman/Cargo.toml: The tinytools Git dependency revision changed from 360721bc3c7a26221b61f6629f5f0415cfccb568 to 1cd9e9c4a117336e6f42d0c7936ed60e7476d88f.
• observed — Modified behavior in vendor/openhuman: The vendor/openhuman subproject commit reference was updated from 467ab703acccb63391a8dd6a530297ef904dda76 to 1f2e59a238c3221004aabda3e5facc2f5624eaa6.

---

---

Pre-merge checks | 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Docstring Coverage		No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff. Analyzed 0 functions across 0…
Linked Issues check		Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check		Check skipped because no linked issues were found for this pull request.
Description Check		Check skipped - CodeRabbit’s high-level summary is enabled.
Title check		The title clearly identifies the primary change: updating OpenHuman to include prompt-cache and usage fixes. It is concise and specific.

---

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

---

_{A rabbit checks each pinned hash,
Then hops along the meadow path.
New revisions take their place,
The burrow builds at steady pace,
And clover waits beside the stash.}

_{Comment @coderabbitai help to get the list of available commands.}