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.
git clone --depth 1 https://github.com/ViryaZheng/recomby-geoWrote 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/commands/viryazheng/recomby-geo/01-intake)<a href="https://agentmods.dev/commands/viryazheng/recomby-geo/01-intake"><img src="https://agentmods.dev/badge/commands/viryazheng/recomby-geo/01-intake/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/commands/viryazheng/recomby-geo/01-intake"><img src="https://agentmods.dev/badge/commands/viryazheng/recomby-geo/01-intake.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00000 | $0.02050 |
| Opus 5 | $0.00000 | $0.01025 |
| Sonnet 5 | $0.00000 | $0.00410 |
| Haiku 4.5 | $0.00000 | $0.00205 |
Grade A, and why
01-intake 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 11d 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.
How it starts
The opening of the file, as written. The whole thing — 194 lines — stays where its author put it; the contents beside it link to each section on GitHub.
01 · Intake — Build the Brand Context
The single source of truth for the entire pipeline. Every downstream skill
reads brand_context.json. If this is wrong or thin, every later skill
amplifies the error.
Core principle: AI recommends whoever best answers the question. The richer and more precise this profile, the more accurately the rest of the pipeline can identify the exact questions this business should own.
Inputs
clients/<slug>/inputs/— materials provided by the client (any of: PDF, DOCX, PPTX, XLSX, image, plaintext notes, URL list).- The user may also provide context conversationally during this command run.
Output
clients/<slug>/brand_context.json— validates againstschemas/brand_context.schema.json(3 layers + extended).clients/<slug>/intake-log.md— append-only human-readable log of what was extracted from which source.
Procedure
Run sequentially. Do not signal readiness to 02-audit until every gate passes.
Step 1 — Ingest materials
For each file in clients/<slug>/inputs/:
| Input type | Approach |
|---|---|
*.pdf |
Read with the Read tool (built-in PDF support up to 10 pages; for larger PDFs, pass pages ranges) |
*.docx |
unzip -p file.docx word/document.xml | sed 's/<[^>]*>/ /g' then tr -s ' ' '\n' |
*.pptx |
Same as docx but ppt/slides/slide*.xml |
*.xlsx / *.csv |
python3 -c "import pandas; print(pandas.read_excel(...).to_csv())" |
| URL list | WebFetch each URL with extraction prompt |
*.png / *.jpg |
Read (multimodal) — extract text and structural info |
| Notes | Read directly |
Do not dump raw content into brand_context.json. Apply the extraction filter: keep only signals that map to a schema field. Discard the rest.
Step 2 — Layer 1 (Business Identity) — 5 hard-required fields
Fill these from materials. If any is missing or thin (single word, "TBD", generic-quality phrase), ASK the user a targeted question. Don't continue with placeholder values.
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.
- 11d ago First seen · 194 lines · 0 tokens per session scan A 0cbbf853516b
01-intake is a command published in the GitHub repository ViryaZheng/recomby-geo (454 stars, last pushed 2mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 2,050 tokens. 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.
Other commands, from other repositories
commit-all
Stage and commit all current changes with a single message. Auto Git Push will pick up and push to origin within 30s.
check
Run the full local quality gate — TypeScript check, build, dependency audit (prod-only), skill registry validate. Fast pre-commit / pre-suggestdeploy sweep.
registry
Refresh the skills registry — regenerate SHA-256 manifest from disk, then validate. Run after editing any file under .agents/skills/.
onboard
/anty:onboard — QUEST-Based Conversational Interview.
review
/anty:review — 5-Question Review Engine.
scan
/anty:scan — Analysis Cycle (10-Point).