https://memfor.ai/changelog What shipped, with the commit that shipped it. 2026-08-17T00:00:00Z memfor https://memfor.ai/changelog#2026-08-17-first-community-pull-request 2026-08-17T00:00:00Z

An outside contributor added the coding-interview practice skill (#10).

<p>An outside contributor added the coding-interview practice skill (#10).</p><p>Commit <a href="https://github.com/r5rana/agentware/commit/3c25cddd28b5e4d1c320c1e34a98eacaecf6a199"><code>3c25cdd</code></a></p> https://memfor.ai/changelog#2026-07-24-single-pass-adversarial-review 2026-07-24T00:00:00Z

Changes to agentware's own code and steering get one adversarial audit and a self-verified fix before they land.

<p>Changes to agentware&#39;s own code and steering get one adversarial audit and a self-verified fix before they land.</p><p>Commit <a href="https://github.com/r5rana/agentware/commit/0c41fc9105ef35371cf498276bb003f2eed31956"><code>0c41fc9</code></a></p> https://memfor.ai/changelog#2026-07-19-approval-gated-skill-cache 2026-07-19T00:00:00Z

Skills run from a vetted, approval-gated cache, so a skill synced from a team repository doesn't run until it's approved.

<p>Skills run from a vetted, approval-gated cache, so a skill synced from a team repository doesn&#39;t run until it&#39;s approved.</p><p>Commit <a href="https://github.com/r5rana/agentware/commit/c16e24740ecc2c8b6028c34c16c410a5beb3b241"><code>c16e247</code></a></p> https://memfor.ai/changelog#2026-07-12-signed-tag-updater 2026-07-12T00:00:00Z

The package updater only moves to signed release tags. It's off by default.

<p>The package updater only moves to signed release tags. It&#39;s off by default.</p><p>Commit <a href="https://github.com/r5rana/agentware/commit/34bc4b19b21229d42ffcb9899e345f45787f8707"><code>34bc4b1</code></a></p> https://memfor.ai/changelog#2026-06-28-local-models-lm-studio 2026-06-28T00:00:00Z

Run the execute phase on a local model while planning stays on your cloud runtime.

<p>Run the execute phase on a local model while planning stays on your cloud runtime.</p><p>Commit <a href="https://github.com/r5rana/agentware/commit/a04d3a9ebab2699cc7ebcd097031a9713eca2475"><code>a04d3a9</code></a></p> https://memfor.ai/changelog#2026-06-27-semantic-recall-and-codex 2026-06-27T00:00:00Z

Mode B adds a pinned, on-device embedding model on top of BM25. It's opt-in, and BM25 stays the default. OpenAI Codex joins Claude Code as a supported runtime.

<p>Mode B adds a pinned, on-device embedding model on top of BM25. It&#39;s opt-in, and BM25 stays the default. OpenAI Codex joins Claude Code as a supported runtime.</p><p>Commits <a href="https://github.com/r5rana/agentware/commit/938486c1b5de281453752b85dc996148444fffa0"><code>938486c</code></a>, <a href="https://github.com/r5rana/agentware/commit/08157ebb636f9dafcfd5945ab653fcecde4c4837"><code>08157eb</code></a></p> https://memfor.ai/changelog#2026-06-25-apache-licence 2026-06-25T00:00:00Z

agentware gets its LICENSE and NOTICE files.

<p>agentware gets its LICENSE and NOTICE files.</p><p>Commit <a href="https://github.com/r5rana/agentware/commit/60016eefe0686d31622520219775adcdf6fccc78"><code>60016ee</code></a></p> https://memfor.ai/changelog#2026-06-25-longmemeval-scorer 2026-06-25T00:00:00Z

Recall@5 0.9140 on 470 answerable questions, recorded from a clean tree.

<p>Recall@5 0.9140 on 470 answerable questions, recorded from a clean tree. <a href="https://memfor.ai/benchmarks/runs/42d58a0-longmemeval-20260625t153444z">Receipt</a></p><p>Commit <a href="https://github.com/r5rana/agentware/commit/42d58a042c8826d10c298f6a650bd870e89f194d"><code>42d58a0</code></a></p> https://memfor.ai/changelog#2026-06-23-repository-created 2026-06-23T00:00:00Z

The first commit holds only a README.

<p>The first commit holds only a README.</p><p>Commit <a href="https://github.com/r5rana/agentware/commit/19834fb0fae106f9a2f0673208cdc0f5041ef0aa"><code>19834fb</code></a></p>