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/firatcand/forge/delivernpx skills add firatcand/forge --skill delivergit clone --depth 1 https://github.com/firatcand/forgeWrote 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/firatcand/forge/deliver)<a href="https://agentmods.dev/skills/firatcand/forge/deliver"><img src="https://agentmods.dev/badge/skills/firatcand/forge/deliver.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 | $0.00064 | $0.04599 |
| Opus 5 | $0.00032 | $0.02299 |
| Sonnet 5 | $0.00013 | $0.00920 |
| Haiku 4.5 | $0.00006 | $0.00460 |
Grade A, and why
deliver 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 5d 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.
/deliver
/deliver <phase-id | --all> drives a whole band of work — a phase, a
feature-set, or the entire roadmap — toward "merged + Done", by grouping ready
tickets into themed batches and driving each batch through the per-ticket
/drive gauntlet. It is the long-horizon recipe above /drive: /drive
delivers one ticket; /deliver sequences and batches many, front-loads the
architectural decisions, and runs the phase-gate ceremony at phase boundaries.
This is the vision headline — use loops, goals, and the gauntlet to deliver a
full phase/feature-set/product by asking the user mainly architectural
questions — codified into a reusable skill that orchestrates the EXISTING
skills (/drive, /pickup-task, /plan-task, /implement, /review,
/second-opinion, /ship, /wrap-up, /phase-gate) and read verbs.
This is a per-turn recipe — NOT a loop (ADR)
/deliver does all it can in one invocation and then RETURNS. It MUST NOT
self-loop: no sleep, no watch, no while true, no --follow, no
gh pr checks --watch. Whenever the next step is an asynchronous wait — CI
running, a human decision pending — /deliver reports the state and returns.
The repeat-until-done behavior is owned by the native /goal driver:
/goal "phase-4 delivered" # or: /goal "v0.5 delivered"
and the host re-invokes /deliver <band> each turn until every ticket in the
band is merged and Done. Keeping the loop in native /goal//loop (not here) is
what lets every round run as a fresh, fully-budgeted interactive turn — the
autopilot ADR decision (spec/decisions / the native-/goal ADR). /deliver is
the recipe; /goal is the engine. Within one turn /deliver may drive several
batches sequentially, but it never wraps itself in its own loop.
Runs on the interactive skill path (NOT the formal dispatch lifecycle)
/deliver uses the SAME interactive recipe as /drive and /pickup-task:
tracker status (set natively) + ensure-worktree + /plan-task +
/implement + /review + /second-opinion + /ship + tracker-Done +
/wrap-up (whose gc --remove-worktrees removes the worktree). It does NOT
drive the formal claim → dispatch → heartbeat → complete state machine — that
is the HEADLESS-orchestrator path. This matters two ways:
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.
- 5d ago First seen · 335 lines · 64 tokens per session scan A 586a84dddc3b
deliver is a skill published in the GitHub repository firatcand/forge (13 stars, last pushed 1mo ago), licensed MIT. It adds 64 tokens to every session and 4,599 once invoked, about $0.0003 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
ospec-change
Create or advance a lightweight OSpec change using the classic fast workflow.
prospec-knowledge-generate
Generate AI Knowledge - Read raw-scan.md, analyze project structure, autonomously decide module boundaries, and produce Recipe-First module READMEs and index. Triggers: generate knowledge, analyze project, module split, 產生知識, 知識庫, 分析專案, 模組拆分.
submit-pr
Open a prospec pull request in the house format — push the change's two commits, write the Traditional Chinese body, and link it to its issue. Triggers: submit pr, open pr, 開 PR, 送 PR, 發 PR, pull request, 提交 PR.
prospec-explore
Explore - Requirement exploration, problem investigation, and solution comparison partner. Triggers: explore, compare, investigate, unsure, clarify, 探索, 比較, 釐清, 調查, 不確定.
prospec-plan
Plan Implementation - Convert User Story into technical implementation plan (plan.md) and change specification (delta-spec.md). Triggers: plan, architecture, technical plan, 規劃, 架構規劃, 技術規劃.
prospec-knowledge-update
Incremental Knowledge Update - Parse delta-spec.md to identify affected modules, scan source code, and update module README, index.md, and module-map.yaml incrementally. Triggers: knowledge update, incremental update, sync knowledge, update docs, 更新知識, 增量更新, 同步知識, 更新文件.