work

A command for starting or continuing an organized software-development work session.

In plain words
What is it for?
For working on all tasks or selected epics and stories, continuing unfinished work, choosing supervised or automated execution, and setting parallel task limits.
Why use it?
It gives the session a defined operating mode, task scope, level of parallel work, and optional review or automation settings.

Command

Install

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.

agentmods
npx agentmods add commands/srstomp/pokayokay/work
Clone the repo
git clone --depth 1 https://github.com/srstomp/pokayokay
Per session 7 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 15,351 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00007 $0.15351
Opus 5 $0.00003 $0.07676
Sonnet 5 $0.00001 $0.03070
Haiku 4.5 $0.00001 $0.01535

Measured 2d ago against content hash fa82e597bb59, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

work 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.

plugins/pokayokay/commands/work.md · 1,706 lines

How it starts

The opening of the file, as written. The whole thing — 1,706 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Work Session Workflow

Start or continue a development session with configurable human control.

Mode: $ARGUMENTS (default: supervised) Parallel: Extract -n N or --parallel N from arguments (default: 1, "auto" for adaptive)

Argument Parsing

Parse $ARGUMENTS to extract:

  1. Mode: First word if it matches supervised|semi-auto|auto|unattended, else "supervised"
  2. Parallel: Value after -n or --parallel flag. Values: auto, or 1-5. Default: 1
  3. Scope: --epic <id>, --story <id>, or --all (limits which tasks to work on)
  4. Continue: --continue flag (resume from previous session's WIP)
  5. Skip design: --skip-design flag (bypass the Design Review Gate for all tasks this session)
  6. Skip brainstorm: --skip-brainstorm flag (skip the Brainstorm Gate for all tasks this session)
  7. Skip auto-fix: --skip-auto-fix flag (disable auto-fixer dispatch on test failures for this session)

Example arguments:

  • semi-auto -n 3 → mode=semi-auto, parallel=3 (fixed)
  • semi-auto -n auto → mode=semi-auto, parallel=adaptive (starts at 2)
  • --parallel 2 → mode=supervised, parallel=2 (fixed)
  • auto → mode=auto, parallel=1
  • unattended → mode=unattended, parallel=1 (never pauses, for overnight runs)
  • semi-auto --epic epic-abc123 → mode=semi-auto, scope=epic:epic-abc123
  • auto --story story-def456 -n 3 → scope=story:story-def456, parallel=3
  • --continue → resume from WIP, inherit previous scope
  • --all → work on all available tasks (no scope filter)
  • auto --skip-design → mode=auto, design review gate bypassed

Note: -p is commonly reserved by AI runtime CLIs for prompt input. Use -n for parallel count.

Worktree Argument Parsing

Extract worktree flags from $ARGUMENTS:

  1. --worktree: Force worktree creation (even for chores)
  2. --in-place: Force in-place work (skip worktree)

Example arguments:

  • semi-auto --worktree → mode=semi-auto, forceWorktree=true
  • --in-place → mode=supervised, forceInPlace=true
  • auto -n 3 → mode=auto, parallel=3, useSmartDefault=true

Read the full file on GitHub · 1,706 lines

Changes

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.

  1. 2d ago First seen · 1,706 lines · 7 tokens per session scan A fa82e597bb59

Subscribe to this mod's changes

work is a command published in the GitHub repository srstomp/pokayokay (9 stars, last pushed 1mo ago), licensed MIT. It adds 7 tokens to every session and 15,351 once invoked, about $0.0000 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.