In early October 2026 I released several open‑source tools that together enable fast, reliable AI‑agent workflows, including a persistent knowledge‑graph codebase indexer, a harness‑engineering curriculum, a semantic search CLI, an adaptive harness framework, and a voice‑isolation front‑end.
codebase‑memory‑mcp builds a persistent knowledge graph of my entire codebase that my agent can query in under <1 ms>.
Five structural queries run at around ≈3,400 tokens compared to 412,000 for file‑by‑file grep exploration.
Hybrid LSP semantic type resolution for Python, TypeScript, JavaScript, PHP, C#, Go, C, C++, Java, Kotlin, Rust, and Perl, building call graphs that understand inheritance, generics, and cross‑file imports.
Auto‑sync via a background watcher that detects file changes and re‑indexes automatically.
Team‑shared graph artifact via a compressed SQLite snapshot you commit to the repo so teammates skip the reindex.
On the 10‑task SWE‑bench, both Jevgrep and the baseline solved 8/10 tasks; coding‑agent cost dropped from $7.62 to $5.44 (excluding Jev cost).
MemoHarness decomposes the harness into six editable control dimensions and treats execution history as a learning signal.
Across 265 real recordings and 11 STT configurations, pooled word error rate dropped from 23.3% to 6.2% after Voice Isolation.