Synthorai (https://synthorai.io) reaches Anthropic, OpenAI, Google, Z.ai,
Moonshot, DeepSeek and Qwen models through one OpenAI-compatible base URL,
billed at each upstream's own list price.
- Upgrade SvelteKit to 2.70.2 for framework maintenance updates
- Upgrade Shiki to 4.4.1 across syntax highlighting packages
- Upgrade Vite to 8.2.0 and Rolldown to 1.2.1
- Synchronize npm and pnpm lockfiles with transitive dependencies
- Process PDF attachments through reusable pdf-inspector WASM worker
- Transfer ArrayBuffers and initialize WASM once per worker
- Reject pending conversions and recreate workers after crashes
- Parse attachments without triggering duplicate chat requests
- Surface OCR, encoding, and empty-content conversion errors
- Remove PDF.js pipeline and obsolete transitive dependencies
- Add tests for conversion results and chat boundaries
Prettier reflows these three files, which the branch adds or rewrites. The count
of files that prettier reports goes from 333 to 330. The rest of that count is
work that this branch does not touch.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
- Add PKM YAML metadata guidance to pattern explanations
- Classify frontmatter generation under conversion, extraction, and writing
- Register descriptions and extracts for pattern suggestion workflows
- Credit both contributors in the incoming changelog entry
PR #2092 added Grok (xAI) web search support and updated the go-flags
struct tag in internal/cli/flags.go, but `fabric --help` renders flag
descriptions from the i18n message catalog (internal/cli/help.go maps
"search" -> "enable_web_search_tool"), not from the struct tag. The
catalog value still listed only Anthropic, OpenAI, and Gemini, so the
Grok support stayed undiscoverable in --help even though it works.
Append "Grok" to the enable_web_search_tool value in all 11 locales,
using each locale's own list separator, and update the generated README
help block to match. No functional change.
The zsh and fish completion strings were already corrected in #2179,
so this carries forward only the remaining parts of #2140.
Co-Authored-By: Ken Hartman <khartman@users.noreply.github.com>
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
- Specify input and output tokens in flag help text
- Synchronize metadata descriptions across shell completion definitions
- Localize clarified metadata guidance across supported languages
- Record description clarification in incoming changelog entry
- Declare Fish completion arguments for dynamic and fixed values
- Enable filtered file suggestions for supported path options
- Add Spotify, transcription, metadata, and wire-debug completion options
- Include Grok among providers supporting web search completion
- Localize pattern, Spotify, and metadata help descriptions consistently
- Document expanded completion coverage and wire-level debug behavior
- Describe YouTube visual flags directly in generated help
Adds scripts/audit-patterns.sh, a dependency-free report over the
patterns library and a few adjacent consistency checks:
1. Thin patterns — system.md under 15 lines (likely stubs)
2. Bloated patterns — system.md over 50 KB (likely embedded examples)
3. Stale model refs — hardcoded GPT-4 / ChatGPT / vendor model names
4. Missing INPUT — pattern has no `# INPUT` section
5. i18n key gaps — keys in en.json missing from other locales
6. Completion gaps — long flags in flags.go absent from completions
Reports only; it never edits patterns. Exits 0 by default so it can be
run casually, or `--strict` to exit 1 on findings for CI use.
Against the current tree it reports 109 issues, which is what motivated
the separate normalization and de-bloating changes.
- Translate Bedrock setup prompts across nine supported locales
- Localize datetime and system template errors consistently
- Translate Persian Spotify errors and setup guidance
- Correct Japanese and Polish file operation log labels
