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/tasksnpx skills add get-bb/bb --skill tasksgit 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 | $0.00046 | $0.01230 |
| Opus 5 | $0.00023 | $0.00615 |
| Sonnet 5 | $0.00009 | $0.00246 |
| Haiku 4.5 | $0.00005 | $0.00123 |
Grade A, and why
tasks 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 yesterday.
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 — 137 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Tasks
Use the bb tasks CLI to understand the assigned task, keep its record useful,
and report the outcome where the work is tracked.
Delegation presets are user-defined; Tasks ships with none. Before dispatching
work, use bb tasks preset list and create a preset if the required one does
not already exist. Dispatch requires an existing preset.
Create or update the same execution selection exposed in the Tasks UI with
--provider, --model, --reasoning, and optional
--service-tier default|fast|none:
bb tasks preset create --name "Codex high" --provider codex \
--model gpt-5.6-sol --reasoning high --service-tier fast \
--permission auto
preset update accepts the same flags; --service-tier none clears a tier.
Work a task
-
Find and read the task before acting:
bb tasks show ABC-12The detail includes the description, status, priority, labels, subtasks, comments, attachments, attached worker threads, and the GitHub pull requests those threads produced (from environment metadata, with state open/draft/merged/closed). Use
bb tasks show ABC-12 --jsonwhen the result will drive commands or code.For project-wide discovery,
bb tasks listreturns at most 100 rows by default. Pass--limit 1-500; in JSON, continue withnextCursorvia the same filters/sort and--cursor <value>. A task-list mutation makes an old cursor stale, so restart without it. -
Fetch every relevant attachment before making assumptions about it:
bb tasks attachment get <attachment-id> --out <path> -
Do the work. Post one substantive comment at each meaningful milestone, such as a completed investigation, an implementation ready for validation, or a concrete blocker:
bb tasks comment ABC-12 --body "Implemented the change; focused validation now passes."Add
--notifyonly when the new comment should be delivered to the thread that authored the task's most recent agent reply. This resumes an idle recipient; with no prior agent reply, the comment is recorded without targeting an unrelated thread. In agent context, the new comment keeps the current thread identity and an explicit--author, while delivery still targets the prior latest responder rather than the new comment itself.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- yesterday First seen · 137 lines · 46 tokens per session scan A f6ca58a82d1d
tasks is a skill published in the GitHub repository get-bb/bb (2,762 stars, last pushed 2d ago), licensed MIT. It adds 46 tokens to every session and 1,230 once invoked, about $0.0002 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
emil-design-eng
This skill encodes Emil Kowalski's philosophy on UI polish, component design, animation decisions, and the invisible details that make software feel great.
cradle-cli
Use whenever you need to read, inspect, wait on, or mutate Cradle-owned state or workflows, including issues, delegation, sessions, awaits, CI/review waits, timed waits, Work inspection, workspaces and git inspection, Chronicle, automations, usage, observability, skills, agents, profiles, providers, preferences…
elysiajs
Create backend with ElysiaJS, a type-safe, high-performance framework.
transitions-dev
Production-ready CSS transitions for web apps. Use when implementing notification badges, dropdowns, modals, panel reveals, page transitions, card resizes, number pop-ins, text swaps, icon swaps, success checks, avatar group hovers, or error state shakes. Triggers on "add a transition", "animate the dropdown", "make…
userinterface-wiki
UI/UX best practices for web interfaces. Use when reviewing animations, CSS, audio, typography, UX patterns, prefetching, or icon implementations. Covers 11 categories from animation principles to typography. Outputs file:line findings.
cradle-observability-debugger
Debug Cradle local observability data by querying SQLite events/incidents/timeline, runtime snapshots, metrics, and server logs.