apresmoi
UserAgent skills for driving paid AI subscriptions as external agents, no API keys. Installable as a Claude Code, Codex, and Antigravity plugin.
Categories
Indexed Skills (22)
arrow-diagram
Render ASCII arrow/flow diagrams (fan-out, fan-in, decision trees, retry loops) from a small JSON spec — use whenever an answer needs an arrow-based flow or topology diagram instead of hand-aligning box characters. Triggers on "arrow diagram", "ascii diagram", "flow diagram", "draw the pipeline", "conversation flow".
colab-harness
Use a Google Colab GPU runtime (L4/T4, Colab Pro) as a job server from the local machine through a cloudflared tunnel, no browser automation — faster-whisper transcription, pyannote diarization, YouTube audio download, a vLLM OpenAI-compatible endpoint, LoRA training and DSPy compiles as script jobs. Use when the user asks to run something on Colab, transcribe or diarize audio on a GPU, start vLLM on Colab, train a small model on Colab, or send a batch job to Colab and fetch the results.
deep-research
Run any deep-research prompt through a logged-in ChatGPT Pro or Grok (SuperGrok) session in a real Chrome, wait for the report (15–60 min), and capture it into an intake file — plus research recipes (spread timelines on X, claim checks, origin attribution) and your own named recipes that render a tuned prompt on demand. Use when the user asks to run a ChatGPT/Grok deep-research session, fill a research intake, trace how a story spread on X, or run one of the user's named research recipes. Drives an already-paid subscription seat instead of spending API tokens.
harness-coding-agents
Harness the locally installed subscription coding CLIs — Antigravity (`agy`), Grok (`grok`), Codex (`codex`), and Claude Code (`claude`) — as bounded external agents for analysis, review, alternative solutions, adversarial checks, or explicitly authorized implementation. Use when the user asks to harness grok, run agy, use codex, harness Claude Code, consult Antigravity, get a second opinion or model council, delegate a work packet to an external agent. These CLIs drive paid SUBSCRIPTION logins (no API keys).
colab-video-pipeline
Use this skill when running or maintaining the Jiang Lens Google Colab video pipeline for YouTube download, diarization, transcription, Drive sync, or Playwright-based Colab automation. Requires the project Drive folder named jianglens and never commits cookies, browser profiles, tokens, or downloaded media.
jiang-agent-transcript-pass
Use this skill when an agent processes Jiang Lens transcript packets into semantic JSON outputs: interactions, speaker uncertainty, claims, glossary candidates, and chronology notes.
jiang-autonomous-work-protocol
Use this skill when a Jiang Lens autonomous agent needs to discover, claim, perform, validate, and hand off repo work without overwriting other agents or confusing authored, raw, and generated artifacts.
jiang-corpus-impact-pass
Use this skill after every published Jiang Lens episode, or when reassessing an existing episode, to decide and record how that source mutates the corpus, existing lens pages, lens points, atlas structure, chronology, and canon candidates.
jiang-episode-quality-review
Use this skill when reviewing a Jiang Lens episode or interview PR for public-read quality against the clean transcript, semantic artifacts, source refs, real questions, marks, and website readiness before auto-merge.
jiang-episode-read-writer
Use this skill when writing or revising the public Jiang Lens episode/interview read JSON for one already-ingested video, using its clean transcript and semantic bundle to create a compressed reader-facing public page.
jiang-glossary-entry-writer
Use this skill when creating or revising a Jiang Lens glossary entry for a search-facing term, alias, person, institution, phrase, or query target from dated Jiang source refs and existing lens pages.
jiang-lens-atlas-maintainer
Use this skill when maintaining the public Jiang Lens atlas page, adding or revising lens families, concept clusters, relationships, and stable lens-point anchors from already-published source-grounded material.
jiang-lens-concept-writer
Use this skill when creating or improving one public Jiang Lens concept page from dated source refs, episode reads, semantic bundles, proposals, and existing atlas links.
jiang-lens-distillation
Use this skill as the integration map for Jiang Lens work after episode publication, selecting the narrower corpus impact, concept writing, atlas maintenance, provenance linking, judging, or canon promotion skill needed for the task.
jiang-lens-judge
Use this skill to review Jiang Lens concept docs, episode reads, glossary drafts, canon proposals, or source-grounded public pages with adversarial reader/world-model and grounding/provenance passes before handoff or promotion.
jiang-provenance-linker
Use this skill when adding, repairing, or validating Jiang Lens evidence marks, source refs, lens-point anchors, episode-to-lens links, hover payloads, and backlink surfaces across Markdown and episode read JSON.
jiang-source-ingest
Use this skill when importing synced Jiang Lens source artifacts, resolving YouTube metadata dates, creating canonical source transcripts, and preparing transcript packets for semantic agents.
jiang-transcript-boundary-review
Use this skill when reviewing Jiang Lens transcript boundary candidates from ops/scripts/repair-transcript-boundaries.mjs and producing approve/reject decisions without rewriting the transcript or loading the full transcript into context.
jiang-video-e2e
Use this skill as the integration map for turning one already-transcribed Jiang Lens video from synced Drive artifacts into a website-visible episode or interview, delegating detailed work to the narrower ingest, transcript, read-writing, and publishing skills.
moltnet
Use Jiang Lens Moltnet through the local moltnet CLI: read episode-floor on request, send explicit room messages, and respect each participant's configured wake policy.
jiang-canon-promotion
Use this skill when promoting Jiang Lens proposals, glossary entries, model claims, ledger items, or chronology-sensitive material into canonical content after source grounding and review records exist.
jiang-episode-publisher
Use this skill when taking an already-ingested Jiang Lens source and an existing episode/interview read JSON through generated website data, route validation, transcript/video linkage, and public-site build checks.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.