AugmentClaude
DocsNotesWriting

Conference Report Builder

By KerberosClaw· KerberosClaw/kc_ai_skills· 0

Turn conference recordings and slides into reconstructed sessions and custom reports.

Installation

  1. Make sure Claude is on your device and in your terminal.

    Skills load from ~/.claude/skills/ when Claude Code starts up — so you need it on your machine first. If you don't have it yet, install it once with the command below, then run claude in any terminal to verify.

    One-time setup
    npm i -g @anthropic-ai/claude-code

    Already have it? Skip ahead.

  2. Paste into Claude Code or into your terminal.
    Install
    Sign up to copy
    Sign up free to reveal

    This copies the whole skill folder into ~/.claude/skills/conference-report-kerberosclaw/ — the SKILL.md plus any scripts, reference docs, or templates the skill ships with. Safe default: works for every skill.

    Faster alternative (instruction-only skills)

    Skips the clone and grabs only the SKILL.md file. Don't use this if the skill ships Python scripts, reference markdowns, or asset templates — they won't be downloaded and the skill will fail when it tries to load them.

    Quick install (SKILL.md only)
    Sign up to copy
    Sign up free to reveal
  3. Restart Claude Code.

    Quit and reopen Claude Code (or any other agent that loads from ~/.claude/skills/). New skills are picked up on startup.

  4. Just ask Claude.

    Skills auto-activate when your request matches the skill's description — no slash command needed. Trigger phrases live in the skill's own frontmatter; you can read them in the “What this skill does” section above.

Prefer to read the source first? Open on GitHub.

When Claude uses it

Use when the user attended one or more sessions of a conference (with audio recordings + slide photos) and needs help building (1) faithful per-session reconstructions in markdown (slide visuals + speaker transcript with Whisper hallucination annotations), and (2) a downstream report deliverable whose scope and format are decided interactively with the user. Pipeline phase (raw → mlx_whisper Chinese SRT → per-slide multimodal reconstruction → official agenda cross-check via Playwright if available) is deterministic. Report phase is interactive — always quiz user on scope (single-session / single-day / multi-day synthesis), format (existing template / free-form fallback), recipient (formal / informal), and any business workstream mapping before drafting.

What this skill does

This skill turns conference recordings and slide photos into faithful per-session reconstructions and custom reports. It operates in two distinct phases: a deterministic reconstruction pipeline (Phase A) that produces markdown files combining slide visuals, speaker transcripts with Whisper hallucination annotations, and official agenda cross-checks; and an interactive report-building phase (Phase B) that shapes those reconstructions into deliverables tailored to the user's scope, format, recipient, and business needs.

The pipeline runs on your local Mac using mlx_whisper for transcription, Claude vision for reading HEIC slide photos directly, and optionally Playwright for live agenda lookups. Phase A always follows the same workflow; Phase B begins with a requirements quiz to clarify what you actually need before drafting.

What it does

  • Transcribes audio using mlx_whisper large-v3-turbo with Chinese language detection, outputting SRT files that pair with slide photos
  • Reconstructs slides multimodally by reading HEIC photos directly and extracting titles, bullet points, visuals, layouts, icons, and meaning-carrying elements
  • Annotates Whisper errors inline (proper-noun mishearings, hallucination loops, untranslatable segments) with exact timestamps and severity notes
  • Cross-checks against official conference agendas using Playwright MCP (or curl fallback) to correct speaker names, titles, companies, locations—treating official records as canonical for identification facts
  • Handles edge cases systematically: multi-session audio files (with user-guided splitting), missing recordings, gaps in slide photos, and obscured text (retrieved from public news sources when needed)
  • Produces interactive Phase B reports after a 4-question quiz covering scope (single session / day / multi-day), format (template or default), recipient (formal / informal / personal), and business mapping

How to use it

Invoke the skill with /conf-report <conf-dir-name> [<day-tag>] (e.g., /conf-report cybersec2026 D2). Drop your session materials into drafts/<conf>/raw/<day>_<HHMM-HHMM>/ directories—one folder per session, containing the m4a audio file and HEIC slide photos together. The skill inventories what you have, confirms structure with you, then runs Phase A deterministically to produce per-session markdown files. When you're ready for a report, ask explicitly; Phase B always begins with the requirements quiz to clarify scope, format, recipient, and workstream mapping before drafting.

Note: this skill's internal instructions are written in Chinese. Claude reads them natively and will work with you in English.

Related skills

C

Claude API Helper

anthropics

Build, debug, and optimize Claude API applications with caching and model migration support.

OfficialComplete terms in LICENSE.txt
D

Documentation Co-Authoring

anthropics

Guide structured workflows for writing docs, proposals, and technical specs collaboratively.

Official
P

PPTX Text Extractor

axoviq-ai

Extract text and speaker notes from PowerPoint presentations.

AGPL-3.0-or-later
M

Memory Search

davila7

Search conversation history and recall previous discussions, decisions, and context.