AI work that doesn't evaporate. A workspace for building, tuning, and publishing governed AI products.
This repository also configures its own agents. See what GuideAnts tells them →
Skill Claude CodeCodex
Transcription beyond the GuideAnts wrapper contract: transcribe workspace files by path (no upload, no 50 MB gateway cap), pass language hints, and sideload other qwen3-family ASR snapshots from Hugging Face through the wrapper. Use when a transcription needs a language hint, the file is large, or the user wants to…
Skill Claude CodeCodex
Run TTS model families audio.cpp supports but GuideAnts does not ship — Qwen3 CustomVoice builtin speakers, VibeVoice multi-speaker dialogue, MioTTS, VoxCPM2, PocketTTS, Vevo2 voice conversion — by downloading the model and spawning a private audiocppserver in the sandbox. Use when the user names one of these models…
Skill Claude CodeCodex
Speaker diarization: figure out who spoke when in a recording and produce a speaker-labeled transcript, by spawning a private audio.cpp engine with the sortformer diarization model and labeling each turn with the local ASR engine. Use when the user asks to diarize a meeting or call recording, separate or identify…
Skill Claude CodeCodex
Experimental: use audio.cpp scenarios and models that GuideAnts does not ship — speaker diarization (who spoke when, speaker-labeled transcripts), voice cloning from user-supplied audio, deterministic seeds, builtin speakers, language forcing, and deferred TTS families (Qwen3 CustomVoice, VibeVoice, MioTTS, VoxCPM2…
Skill Claude CodeCodex
Synthesize speech against the user's own host-native audiocppserver build (outside the GuideAnts container) — including model families the container binary lacks, like Kokoro or Parakeet forks. TTS only. Use when the user says they run their own audio.cpp server or asks for a family the container build cannot load.
Skill Claude CodeCodex
Advanced synthesis controls on the loaded audio.cpp TTS model: deterministic output via seed, forcing the spoken language, voice-design from a text description (instructions), and enumerating builtin speaker ids — none of which the built-in GuideAnts audio tools expose. Use when the user wants reproducible audio, a…
Skill Claude CodeCodex
Clone a voice from a user-supplied reference clip and synthesize speech with it, by calling the raw audio.cpp TTS engine with voiceref/referencetext — fields the built-in GuideAnts audio tools do not expose. Use when the user wants generated speech to sound like their own voice or another consenting speaker's…
Skill Claude CodeCodex
Transcription beyond the GuideAnts wrapper contract: language hints and workspace files via the GPU host raw audiocpp gateway (/asr + /files) from a PC sandbox. Use when a transcription needs a language hint or the built-in ASR tool is too narrow.
Skill Claude CodeCodex
Experimental: use full raw audiocppserver scenarios GuideAnts does not ship — speaker diarization, voice cloning, seeds, language forcing, voice-design, and deferred TTS families — via the GPU host token-gated raw gateway (/asr|/tts|/private) from a PC sandbox (or co-located engines if present). Use when the user asks…
Skill Claude CodeCodex
Web and image search through the stack's self-hosted SearXNG (readweb-searxng container): ranked text results with engine attribution, and image search with direct downloads of photos into the notebook. Use when the user wants to search the web, find a photo of something, or get source URLs - no API key and no network…
Skill Claude CodeCodex
Build export-ready PowerPoint decks with clear outlines and speaker notes.
Skill Claude CodeCodex
One-line description for skill discovery. Use when …
Skill Claude CodeCodex
Industry-standard time-coded transcripts (SRT/WebVTT/JSON) via GPU host AudioCPP: product ASR for text + private qwen3forcedaligner (/v1/tasks/run) for word times, collapsed to caption cues. Use when the user wants timed captions without PyTorch.
Skill Claude CodeCodex
Run PowerShell on Windows hosts or a POSIX shell on macOS/Linux hosts over SSH, via the machine registry in the guide Environment (GASSHMACHINES). Use when the user needs host OS tasks — drives, dotnet, Playwright, processes, SSH into their laptop or desktop — not file I/O on paths already visible in the workspace…
Skill Claude CodeCodex
Create and verify single-file interactive HTML pages in the sandbox: build or edit the file, then prove the result with headless-Chromium screenshots, in-page DOM/geometry assertions, and JS-error capture (Playwright). No GPU, no service, no network. Use when the user wants an HTML page, animation, demo, dashboard, or…
Skill Claude CodeCodex
Query the stack's SQL Server (mssql-express) from the sandbox: probe, databases, tables, schema, sample rows, read-only SQL, and opt-in DML/DDL. Connection string comes from GADBCONNECTIONSTRING in the Guide's Environment variables. Use when the user wants to inspect or query the GuideAnts database.
Skill Claude CodeCodex
BF16 image edit via GPU host ComfyUI-video adapter (qwen-image-edit-bf16-v1). Use when the user has a source PNG and a prompt to restyle or complete the scene — not product SD img2img.
Skill Claude CodeCodex
BF16 text-to-image via GPU host ComfyUI-video adapter (qwen-image-v1). Use when the user needs a new PNG from a prompt with Qwen Image 2512 — not product SD txt2img.
Skill Claude CodeCodex
BF16 inpaint via GPU host ComfyUI-video adapter (qwen-image-edit-bf16-inpaint-v1). Use when the user has source + mask PNGs and a fill prompt — masked completion, whiteboard erase, etc.
Skill Claude CodeCodex
Probe GPU host BF16 Qwen Image readiness via the skill gateway and route to generate, edit, or inpaint skills. Use when the user needs Comfy/Qwen workflows beyond product SD tools, or when readiness is unclear.
Skill Claude CodeCodex
Make a talking-head MP4 on Max: avatar + audio + plate → 416×256 LongCat-Video-Avatar-1.5, CorridorKey @ 416×234, BasicVSR++ FG, 1280×720. Submit with videotool.py, then poll status on later sandbox calls.
Skill Claude CodeCodex
Convert a GuideAnts / Code-Executor markdown conversation transcript into a styled, print-ready PDF with a cover page, role-badged turns, full tool-call inputs (heredoc-aware) and highlighted code. Use when the user wants to export, archive, or share a conversation log as a polished PDF document.
Skill Claude CodeCodex
Turn a recorded browser-session capture (screen + mic archive) into a context report, a timed transcript of the narration, and a TTS-ready script of the story the user told. Use when the user records a session and wants narration/script work for a video built from it.
Skill Claude CodeCodex
Turn a final TTS audio track (with its timed transcript) into an editable story timeline (xlsx) for a human to fill with images, slides, and video clips. Audits the audio against timing best practices first, then maps each finding to a visual offset, and QA-scores the timeline objectively. Use when a finished…
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: