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 agentmods add skills/get-bb/bb/skill-creatornpx skills add get-bb/bb --skill skill-creatorgit clone --depth 1 https://github.com/get-bb/bbWhat 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 | $0.00054 | $0.00924 |
| Opus 5 | $0.00027 | $0.00462 |
| Sonnet 5 | $0.00011 | $0.00185 |
| Haiku 4.5 | $0.00005 | $0.00092 |
Grade A, and why
skill-creator 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 2d 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 — 122 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Skill creator
Create skills that trigger for the correct requests and improve agent results.
Workflow
- Read the conversation and the existing skill before asking questions.
- Confirm the trigger, expected result, inputs, edge cases, and dependencies.
- Write or revise the skill and any required resources.
- Test realistic prompts in fresh BB threads.
- Compare the results with the previous skill or no-skill baseline.
- Revise until further changes do not give a useful improvement.
Ask only for decisions that the available context cannot answer. Match the amount of evaluation to the risk and the result quality that the user needs.
Skill contract
A skill uses this structure:
skill-name/
├── SKILL.md
├── references/
├── scripts/
└── assets/
Only SKILL.md is required. User skills live at
~/.bb/skills/<name>/SKILL.md.
The file starts with YAML frontmatter:
---
name: skill-name
description: State what the skill does and when an agent must use it.
---
The directory and name value must match. A name uses lowercase letters,
numbers, and single hyphens. It cannot exceed 64 characters. The description
cannot exceed 1024 characters.
Put all trigger conditions in the description. Include concrete tasks and nearby user phrases. Keep procedural instructions in the body.
BB discovers edits for newly spawned threads. An existing thread does not receive a skill revision after it starts.
Progressive disclosure
BB loads skill information in three levels:
- BB always loads the name and description.
- An agent reads
SKILL.mdafter the skill triggers. - The agent reads or runs bundled resources only when the task needs them.
Keep SKILL.md focused on routing, the main workflow, safety, and success
criteria. Aim for fewer than 500 lines.
Move detailed variants, contracts, and examples into references/. Tell the
agent exactly when to read each reference. Add a table of contents to reference
files longer than 300 lines.
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.
- 2d ago First seen · 122 lines · 54 tokens per session scan A 0c8547ffa0ea
skill-creator is a skill published in the GitHub repository get-bb/bb (2,888 stars, last pushed today), licensed MIT. It adds 54 tokens to every session and 924 once invoked, about $0.0003 per session on Opus 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.
Other skills, from other repositories
vendo-setup
Install and configure Vendo (the embedded product agent) in a host repo. Use when asked to add Vendo to an app, run vendo init/doctor/sync, wire the Vendo handler or VendoProvider, or debug a Vendo install.
promptscript
PromptScript language expert for reading, writing, modifying, and troubleshooting .prs files. Use when working with PromptScript syntax, creating or editing .prs files, adding blocks like @identity, @standards, @restrictions, @shortcuts, @skills, or @agents, configuring promptscript.yaml, resolving compilation errors…
termio
See and drive the sibling agent sessions running alongside you in this Termio project via the termio sessions CLI — list and watch their status, spawn new agent or plain-terminal sessions, send a prompt or an answer into a session, and read an agent's reply from its transcript. Use when delegating work to another…
termio
See and drive the sibling agent sessions running alongside you on this machine via the termiod CLI — list them and their status, start a new agent or plain-terminal session, send a prompt or an answer into one, and set your own status. Use when delegating work to another session, or checking on what other sessions are…
add-llm-model
Use when adding a new LLM/model to stagewise model catalogs, provider instances, routing, validation, docs, or showcase UI.
history-compression
How stagewise's agent history compression pipeline works — boundary selection, recency bias, chained compressions, and the SQLite-backed test harness for replaying real compressions in LLM playgrounds. Use when debugging, tuning, or extending history compression, when investigating context-window overflow, or when the…