Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add korovin-aa97/talkthrough-mcp --skill clawhubgit clone --depth 1 https://github.com/korovin-aa97/talkthrough-mcpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/korovin-aa97/talkthrough-mcp/clawhub)<a href="https://agentmods.dev/skills/korovin-aa97/talkthrough-mcp/clawhub"><img src="https://agentmods.dev/badge/skills/korovin-aa97/talkthrough-mcp/clawhub/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/korovin-aa97/talkthrough-mcp/clawhub"><img src="https://agentmods.dev/badge/skills/korovin-aa97/talkthrough-mcp/clawhub.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00060 | $0.00471 |
| Opus 5.5 | $0.00024 | $0.00188 |
| Sonnet 5.5 | $0.00012 | $0.00094 |
| Haiku 4.5 | $0.00006 | $0.00047 |
Grade A, and why
talkthrough scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
talkthrough for OpenClaw
This skill wires the talkthrough MCP server into OpenClaw and teaches the workflow. Everything runs locally: recordings never leave the machine.
Setup (once)
Add the MCP server:
openclaw mcp add talkthrough --command uvx --arg --python --arg ">=3.11,<3.14" --arg "talkthrough-mcp[diarization,url]"
Requires uv (https://astral.sh/uv). First processing downloads a whisper
model once (~460 MB for the default small).
Workflow
process_media(path)— idempotent by content hash; returns a compact summary with job_id (re-calls on the same file are instant). For one public video/audio URL useprocess_url(url): the source is downloaded once (the only network step), then everything below is local.get_transcript(job_id)/search(job_id, "<word>")— orient; search covers speech AND on-screen OCR text.get_moment(job_id, t0-2000, t1+2000)— evidence bundle per remark: transcript slice + up to 3 frames + OCR + wall-clock range.extract_frame(job_id, at_ms, crop=...)— exact instant, native resolution, when keyframes miss the moment.list_jobs()— recordings processed earlier remain queryable.
Timestamps: t_ms is video-relative; t_wall is real wall-clock time when
the recording start is known — use it to correlate remarks with logs
(±30 s window). Audio-only files (.m4a/.mp3/…) have transcripts but no
frames; frame tools erroring on them is expected.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2mo ago Changed · +2 lines 90a64035a553
- 2mo ago Changed 1fc14f6abb55
- 2mo ago First seen · 40 lines · 60 tokens per session scan A 0d92a9473ac4
talkthrough is a skill published in the GitHub repository korovin-aa97/talkthrough-mcp (33 stars, last pushed 7d ago), licensed MIT. It adds 60 tokens to every session and 471 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
talkthrough compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| talkthroughkorovin-aa97/talkthrough-mcp | Skill | 33 | 7d ago | MIT | A | 60 | Claude Code, Codex |
| opik-diagnosecomet-ml/opik-mcp | Skill | 225 | 9d ago | Apache-2.0 | A | 187 | Claude Code |
| cortex-explore-memorycdeust/Cortex | Skill | 76 | yesterday | MIT | A | 91 | Claude Code |
| qmclaw-workbenchxi-zhao/OpenQuantum | Skill | 71 | yesterday | MIT | A | 124 | Codex |
| quantum-fermi-hubbard-vqexi-zhao/OpenQuantum | Skill | 71 | yesterday | MIT | A | 67 | Claude Code, Codex |
| quantum-groverxi-zhao/OpenQuantum | Skill | 71 | yesterday | MIT | A | 62 | Claude Code, Codex |
| guppy-programsxi-zhao/OpenQuantum | Skill | 71 | yesterday | MIT | A | 48 | Claude Code, Codex |