Tagvico 3.2.5 checks whether a selected model can really make the app’s tool calls
its_artur1 · reddit · 2026-07-29
Tagvico, a free MIT-licensed Paperless sidecar, now probes whether a model can actually execute the exact tool call the app needs during setup.
The author says merely seeing a model in /models only proves the endpoint responds. In version 3.2.5, setup now sends one forced function call with the app’s strict schema to the exact selected model, and invalidates the result if the URL or credentials change. The probe also uses a catalog cap, deadlines, and concurrency limits, while runtime reads and cited answers can run directly but writes become durable proposals that need approval. New installs start with scans paused.
More from coding & agent
- Relume MCP and Claude Code power a Markdown-first design workflow — michalmalewicz · 2026-07-29
- A research talk compares SWE-bench, CodeClash, and ProgramBench for agentic coding — OfirPress · 2026-07-29
- Kuna debuts as an agent-driven Rust decompiler, ranking near Hex-Rays in structuring — moyix · 2026-07-29
- $100k Bet: AI 'Vibecoding' Can't Match Complex SEO Tools Like Ahrefs — eptwts · 2026-07-29
- An AI-agent startup argues billing should work from facts, not scattered API calls — davidcrawshaw · 2026-07-29
- Most Cursor users still treat it like VS Code, despite 1M users and a $29.3B valuation — nikola_mr64990 · 2026-07-29