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 commands/aleksandarbisevac/claude-plugins/taskgit clone --depth 1 https://github.com/AleksandarBisevac/claude-pluginsWrote 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/aleksandarbisevac/claude-plugins/task)<a href="https://agentmods.dev/commands/aleksandarbisevac/claude-plugins/task"><img src="https://agentmods.dev/badge/commands/aleksandarbisevac/claude-plugins/task.svg" alt="Measured on agentmods" 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 | $0.00149 | $0.05746 |
| Opus 5 | $0.00075 | $0.02873 |
| Sonnet 5 | $0.00030 | $0.01149 |
| Haiku 4.5 | $0.00015 | $0.00575 |
Grade A, and why
task 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 4d 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 — 335 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/audit:task — add a task to the manifest, move one between phases, or close one
$ARGUMENTS: subcommand add followed by a quoted title and any of the
flags in the argument-hint above;
or subcommand scope followed by a task id and any of its flags;
or subcommand move followed by a task id and --to <phaseId>;
or subcommand cancel followed by an id and --reason "<why>";
or subcommand priority, the legacy spelling covered at the end of this file.
Unknown/empty subcommand → print usage and stop.
The flags ARE the interface; the dialogue is the fallback. Every answer the
add and scope sections gather is a flag on scripts/manifest/audit-task.py, so a
caller who already holds the specification passes it and is asked nothing — step 2's
ask only for what's missing is what happens to whatever is left. The hint above used
to advertise --phase alone, and what that cost was measured live: an operator holding
files, gate, risk, tests-mode and description was taken through a round of
AskUserQuestion per value, and every question that did not need asking is another
chance to paraphrase a value the caller had already decided — the defect --reason
had (see The operator's words go in VERBATIM below).
priority is deliberately absent from the argument-hint above while remaining a
working subcommand. The hint is what a reader is offered when they type the command, and
offering the old spelling there is how the old spelling keeps being learned — the
opposite of what an alias is for.
0. Conventions
Read ${CLAUDE_PLUGIN_ROOT}/reference/manifest-conventions.md FIRST. Resolve and read
the manifest. If it doesn't exist, stop and point to /audit:init (or the starter template).
add writes through scripts/manifest/audit-task.py, which takes and releases the index lock
itself; hold the lock by hand (conventions → Concurrency lock) only around writes YOU make
with Edit — which is now the move subcommand and nothing else. Creating a phase is
/audit:phase add and is no longer done here by hand (see step 1).
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.
- 4d ago First seen · 335 lines · 149 tokens per session scan A 227613559cee
task is a command published in the GitHub repository AleksandarBisevac/claude-plugins (4 stars, last pushed 3d ago), licensed MIT. It adds 149 tokens to every session and 5,746 once invoked, about $0.0007 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 2026-08-31.
Other commands, from other repositories
init
在既有專案初始化 Octopus——摸專案、檢查 OpenSpec CLI 與結構（缺則引導安裝）、盤點 change 狀態、建 Arena、出交接報告。每個專案跑一次。.
list
Show the project backlog (open items by default, sorted by priority then age). Ask for --all, --sort date or --priority high and Claude re-runs the engine with them.
prune
Delete old closed entries (status done/cancelled) left over from earlier plugin versions. Shows the list first; the deletion itself asks for your approval.
build
SDD 後半段（全自主）：驗 Locked → Builder 逐 task 實作（每條 task 隨行回報）→ 審查 → P1 自動退修（≤3 輪）→ 帶驗收報告回來，你只剩 merge 與 archive。中途不停下來等人。.
add
Add a task to the backlog manually. Use when the user explicitly asks to record something in the backlog.
cancel
Cancel a backlog task by id (the entry is deleted from the backlog).