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 skills add Jamie-BitFlight/claude_skills --skill work-milestonegit clone --depth 1 https://github.com/Jamie-BitFlight/claude_skillsWrote 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/skills/jamie-bitflight/claude_skills/work-milestone)<a href="https://agentmods.dev/skills/jamie-bitflight/claude_skills/work-milestone"><img src="https://agentmods.dev/badge/skills/jamie-bitflight/claude_skills/work-milestone/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/jamie-bitflight/claude_skills/work-milestone"><img src="https://agentmods.dev/badge/skills/jamie-bitflight/claude_skills/work-milestone.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00123 | $0.04098 |
| Opus 5.5 | $0.00049 | $0.01639 |
| Sonnet 5.5 | $0.00025 | $0.00820 |
| Haiku 4.5 | $0.00012 | $0.00410 |
Grade A, and why
work-milestone 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 1mo 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 — 300 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/work-milestone
Execute a groomed milestone. Reads the dispatch plan produced by /groom-milestone, creates an integration branch, spawns parallel kage-bunshin sessions (independent claude -p processes) per wave — each in its own worktree with full orchestrator capabilities (Agent tool, all MCP servers). Sequentially merges their branches and lands the integration branch to main when all waves complete.
Entry Conditions
- Milestone number provided as argument
- Dispatch plan exists and readable via
dispatch_read(milestone_number=N) - All items in dispatch plan are groomed (
groomed: true) - Backlog MCP and SAM MCP responding
- Clean git state on main (if dirty, run
plugins/development-harness/scripts/prepare_clean_worktree.sh {integration-branch}and accept/reject the stash prompt)
Run /groom-milestone {N} first if the dispatch plan is missing or stale.
Main Workflow
flowchart TD
Start(["Input: milestone number N"]) --> LoadPlan["Step 1: Load Dispatch Plan<br>Call dispatch_read(milestone_number=N)<br>Output: waves, conflict groups, quality gates"]
LoadPlan --> Validate{"Step 2: Validate Plan<br>Plan found?<br>All items still open?<br>Observable: dispatch_read result + backlog_list_issues(milestone=N)"}
Validate -->|"Plan missing"| Block1(["BLOCKED — run /groom-milestone {N}"])
Validate -->|"Items changed since groom"| Regroom["Re-run /groom-milestone {N}"]
Validate -->|"Plan valid"| CreateBranch
Regroom --> LoadPlan
CreateBranch["Step 3: Create Integration Branch<br>github_branches create (GitHub backend only)<br>Fallback (beads backend): git checkout -b milestone/{N}-{slug} origin/main<br>Branch: milestone/{N}-{slug} — when backend=beads, N is the beads ID (e.g. bd-a3f8)<br>Switch to integration branch locally.<br>Run prepare_clean_worktree.sh before any worktree add."]
CreateBranch --> FetchItems["Step 3b: Fetch All Items Once<br>Call backlog_view once per issue in the dispatch plan.<br>Store results in context keyed by issue number.<br>DO NOT call backlog_view again for any issue<br>already fetched this session."]
FetchItems --> WaveLoop["Step 4: Wave Dispatch Loop<br>Read next wave from dispatch plan"]
WaveLoop --> WaveItems{"Wave has items?"}
WaveItems -->|"No waves remain"| Land
WaveItems -->|"Yes"| CreateWorktrees["Step 5a: Create Worktrees<br>For each item in wave:<br>git worktree add worktrees/{slug}<br>from integration branch"]
CreateWorktrees --> WriteLocks["Step 5b: Write Lock Files<br>For each worktree:<br>Write .claude/kage-bunshin.lock<br>(prevents recursive spawning)"]
WriteLocks --> SpawnSessions["Step 5c: Spawn Kage-Bunshin Sessions<br>For each worktree: cd into it, then:<br>claude -p --model {model} [--effort {level}]<br>--permission-mode auto --output-format json<br>'Load /dh:work-backlog-item #{issue}'<br>Each session is a FULL orchestrator —<br>has the Agent tool, all MCP.<br>All sessions launch as background processes."]
SpawnSessions --> WaitReturn["Step 6: Wait for All Sessions<br>Poll PIDs for exit.<br>Read result JSON from each<br>/tmp/kb-work-{issue}.json"]
WaitReturn --> CheckResults{"Step 6a: Any sessions failed?"}
CheckResults -->|"All succeeded"| ParseResults
CheckResults -->|"Some failed"| InvestigateFail{"Fixable?"}
InvestigateFail -->|"Yes"| FixRetry["Fix and re-spawn<br>failed item only"]
FixRetry --> WaitReturn
InvestigateFail -->|"No"| SkipFailed["Skip failed item<br>Log to output"]
SkipFailed --> ParseResults
ParseResults["Step 6b: Read Outcomes<br>plan status --plan-address {plan_ref} per item<br>for each task's status and result.<br>Extract BRANCH, FILES_CHANGED, COMMITS, NOTES<br>from the report body in each result JSON."]
ParseResults --> MergeLoop["Step 6b: Merge Worktree Branches<br>Sequential merge into integration branch.<br>One at a time, in return order."]
MergeLoop --> MergeResult{"Any merge conflicts?"}
MergeResult -->|"All clean"| WaveComplete
MergeResult -->|"Trivial/Medium conflict"| Resolve["Auto-resolve or spawn<br>conflict-resolution agent"]
Resolve --> ResolveGates{"Resolution gates pass?"}
ResolveGates -->|"Yes"| WaveComplete
ResolveGates -->|"No"| EscalateConflict["Create backlog item<br>for conflict resolution.<br>Add to milestone."]
MergeResult -->|"Heavy conflict (3+ files)"| EscalateConflict
WaveComplete["Step 6c: Wave Complete<br>All branches merged.<br>Delete worktree branches."]
WaveComplete --> DiscoveryRelay["Step 6d: Discovery Relay<br>Build relay document from agent outputs.<br>Include FILES_CHANGED, COMMITS, NOTES<br>in next wave agent prompts."]
DiscoveryRelay --> PartialCheck{"Any task rows blocked or failed<br>on the ledger?"}
PartialCheck -->|"Yes"| HandlePartial["Create backlog items for<br>blocked tasks. Add to milestone."]
PartialCheck -->|"No"| NextWaveCheck
HandlePartial --> NextWaveCheck{"More waves?"}
NextWaveCheck -->|"Yes"| WaveLoop
NextWaveCheck -->|"No"| Land
Land["Step 8: Land Integration Branch<br>Run full quality gate suite<br>pre_merge + post_merge"]
Land --> FinalResult{"Gates pass?"}
FinalResult -->|"Fail"| FinalFix["Spawn fix agent in worktree<br>on integration branch."]
FinalFix --> Land
FinalResult -->|"Pass"| MergeMain["Step 9: Merge to Main<br>git switch main<br>git merge --no-ff milestone/{N}-{slug}<br>git push origin main"]
MergeMain --> Complete["Step 10: Complete Milestone<br>/complete-milestone {N}.<br>Delete integration branch.<br>Print pending auto-stash reminder(s)."]
Complete --> Done(["Exit: milestone complete"])
What ships with it
2 files 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.
- 1mo ago Changed · +7 lines 3ff3aa93429a
- 1mo ago First seen · 293 lines · 123 tokens per session scan A 5bbd9dfa299e
work-milestone is a skill published in the GitHub repository Jamie-BitFlight/claude_skills (67 stars, last pushed 4d ago), licensed MIT. It adds 123 tokens to every session and 4,098 once invoked, about $0.0005 per session on Opus 5.5. 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 2026-09-14.
Other skills, from other repositories
antfu-issue-triage
Triage open issues and PRs of the current GitHub repo - find already-fixed, duplicate, and spam items, suggest labels, gauge demand from reactions and comments, then present a table of proposed actions and comments for confirmation before touching anything. Use when asked to triage, organize, clean up, or groom issues…
mcore-split-pr
Split a PR into multiple PRs to reduce the number of required CODEOWNERS reviewer groups.
engineer-bot-ops
A workflow for carrying an engineering task from scope through implementation and reviewable delivery. It works with a repository's branch, existing changes, rules, checks, and permissions while separating completed, unverified, and next-step work.
shiplog
Recap of everything shipped since the last run - cross-repo PRs, security fixes, star deltas, and X traction, synthesized into a digest article and a ready-to-post shiplog in your voice.
connect
Choose which wiki this session uses, the team project's pages on the Nimbalyst server or the project's local wiki of files, and connect a repository to its team project. Use at the start of a task before any wiki tool, when a wiki tool returns reponotbound, ambiguousproject, pinmismatch, projectnotaccessible…
llxprt-issue-workflow
Use this skill when asked to address, fix, or work on a GitHub issue in the llxprt-code repository. Covers the complete issue lifecycle - branch setup, gh CLI usage, test-first planning, subagent delegation and review loops, the full verification cycle (including the zai-glm-flash smoke test), open code review (ocr)…