Skip to content

nfemmanuel/iranti

v0.3.15 Breaking

This release includes breaking changes for platform teams planning a safe upgrade.

✓ No known CVEs patched
Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Summary

AI summary

Fix: attend_required now instructs models to retry the blocked operation after successful attendance.

Full changelog

What's changed

Fix: attend_required now tells models to retry the blocked operation

When a discovery tool (search, query, etc.) was blocked with attend_required, the protocol violation message told the model to call iranti_attend(phase='pre-response') — but didn't say to retry the original blocked operation afterward. Models interpreted the successful attend with empty facts as meaning no data existed, rather than as a gate to pass through.

The message now explicitly says: "after attend succeeds, RETRY this exact operation".

Context: This was observed in Copilot CLI where the model called iranti_search, got blocked, called attend, got empty facts, and concluded no memory existed — never retrying the search.

No breaking changes. All 149 contracts pass.

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

Track nfemmanuel/iranti

Get notified when new releases ship.

Sign up free

About nfemmanuel/iranti

Persistent shared memory for AI coding agents. Stores facts as `entity/key/value` triples with hybrid semantic search, task checkpoints, and conflict resolution — shared across Claude Code, Codex CLI, and GitHub Copilot.

All releases →

Beta — feedback welcome: [email protected]