check-work

A read-only start-of-session check for a Casebook work board, a project tracker that records work in progress, upcoming, blocked, and recently completed tasks.

In plain words
What is it for?
Use it to inspect the current branch and uncommitted files, read board cards by status, review recent commits, and identify unlinked or stalled work.
Why use it?
It shows the current repository and task state before work begins, reducing the chance of picking the wrong task or missing recent changes.

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/vldesigns/casebook/check-work
Clone the repo
git clone --depth 1 https://github.com/vldesigns/casebook
Per session 16 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 527 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 $0.00016 $0.00527
Opus 5 $0.00008 $0.00264
Sonnet 5 $0.00003 $0.00105
Haiku 4.5 $0.00002 $0.00053

Measured yesterday against content hash e330ba65a0e0, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

check-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 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.

commands/check-work.md · 47 lines

What it actually says

Give me a clean read of the Casebook board so I know what to pick up. This is the start-of-session bookend to /casebook:end-work.

Read-only orientation pass — no code edits, no board writes, no git operations beyond read-only inspection. If .casebook/state.json doesn't exist, suggest running /casebook:init and stop.

Step 1: Branch + working tree.

  • Run: git status
  • Run: git branch --show-current
  • Report: current branch, clean vs dirty tree, any uncommitted files, ahead/behind vs origin if safe to check.

Step 2: Read the board (the CLI is the source of truth).

  • Run: "$CLAUDE_PLUGIN_ROOT/scripts/tracker" list
  • Report, grouped by status:
    • Today's focus line (verbatim).
    • In Progress cards (title + first line of notes + linked commits + days since last update).
    • Up Next cards (top 3 by most recent update; title + first line of notes).
    • Blocked cards (title + what specifically is being waited on per the notes).
    • Done cards from the last 48h (title + commit hashes).

Step 3: Cross-reference recent commits.

  • Run: git log --pretty=format:'%h %s' -n 10 --author="$(git config user.email)"
  • Flag: any commits not linked to a card; any In Progress card with no linked commits; any In Progress card whose linked commits are already on the main branch (check with git branch --contains ) and may belong in Done.

Step 4: Surface mismatches.

  • If today's focus names work that isn't In Progress or Up Next, flag it.
  • If a Blocked card's notes mention something that may have unblocked (a merged PR, a shipped dependency, an answered question), flag it for review.

Step 5: Propose the next move.

  • Based on today's focus and the In Progress / Up Next state, propose a single concrete next action (one sentence) and which card it maps to.
  • If multiple plausible moves exist, list up to 3 with a one-line tradeoff each.
  • Don't pick for me. Wait for me to direct the work.

Output as a compact status report with simple section headings, not prose. Under 500 words. Wait for me after the report.

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. yesterday First seen · 47 lines · 16 tokens per session scan A e330ba65a0e0

Subscribe to this mod's changes

check-work is a command published in the GitHub repository vldesigns/casebook (1 stars, last pushed 1mo ago), licensed MIT. It adds 16 tokens to every session and 527 once invoked, about $0.0001 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.