Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/xg-gh-25/SwarmAInpx agentmods add skills/xg-gh-25/swarmai/s_deliverWrote 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/xg-gh-25/swarmai/s_deliver)<a href="https://agentmods.dev/skills/xg-gh-25/swarmai/s_deliver"><img src="https://agentmods.dev/badge/skills/xg-gh-25/swarmai/s_deliver/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/xg-gh-25/swarmai/s_deliver"><img src="https://agentmods.dev/badge/skills/xg-gh-25/swarmai/s_deliver.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00091 | $0.01919 |
| Opus 5 | $0.00046 | $0.00959 |
| Sonnet 5 | $0.00018 | $0.00384 |
| Haiku 4.5 | $0.00009 | $0.00192 |
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 11d 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 — 241 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Delivery Packaging
The terminal stage of the lifecycle pipeline. Assembles all artifacts from a pipeline run into a structured deliverable that humans (or future pipeline runs) can review, approve, and act on.
Works at L0 (generates a session summary report). Full artifact bundling at L1+.
What Delivery Produces
| Output | Where It Goes | Who It's For |
|---|---|---|
| Delivery Report (markdown) | Chat + optional Knowledge/Reports/ |
Human review |
| PR Description | Clipboard / chat | Code reviewers |
| Decision Log | Appended to PROJECT.md |
Future sessions |
| Attention Flags | Chat + optional Radar Todo | Human action items |
| Updated PROJECT.md | Projects/<name>/PROJECT.md |
Next session context |
| Updated IMPROVEMENT.md | Projects/<name>/IMPROVEMENT.md |
Learning loop |
| delivery artifact | .artifacts/delivery-*.json |
Pipeline completion marker |
Workflow
Step 1: Gather Artifacts
Collect all artifacts from the current pipeline run:
L0 (no project):
- Scan the current session for: code changes, decisions made, issues found
- No artifacts to read — derive from conversation
L1+ (project with .artifacts/):
- Read
manifest.jsonfor all artifacts in the current pipeline run - Load each active artifact's summary and key data points
- Check pipeline state — delivery should be the terminal state
Artifacts to collect (in pipeline order):
evaluation -> scope, acceptance criteria, ROI score
research -> key findings, sources
alternatives -> chosen approach, rejected approaches + reasons
design_doc -> decisions, API contract, data model
changeset -> files changed, commits, branch
review -> findings, approval status, security issues
test_report -> pass/fail, bugs fixed, remaining issues
Step 2: Assemble Delivery Report
## Delivery Report: <feature/task title>
### Summary
<2-3 sentences: what was built, why, and current status>
### Pipeline Path
EVALUATE -> THINK -> PLAN -> BUILD -> REVIEW -> TEST -> DELIVER
(checkmarks for completed stages, X for skipped)
### What Was Built
- <key change 1: file/component + what changed>
- <key change 2>
- <key change 3>
### Key Decisions
| Decision | Rationale | Alternative Considered |
|----------|-----------|----------------------|
| <decision 1> | <why> | <what was rejected> |
### Quality Summary
- **Tests:** X passed, Y failed, Z fixed during QA
- **Security:** N findings (M auto-fixed, K reported)
- **Review:** <approval status>
### Unresolved Issues
1. <issue + file:line + severity>
2. <issue + diagnosis + suggested fix>
### Attention Flags (Human Review Needed)
- [ ] <item requiring human decision>
- [ ] <item requiring human review>
### Suggested Next Actions
1. <action 1>
2. <action 2>
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.
- 11d ago First seen · 241 lines · 91 tokens per session scan A 9a4a69bff008
deliver is a skill published in the GitHub repository xg-gh-25/SwarmAI (44 stars, last pushed 4d ago), licensed MIT. It adds 91 tokens to every session and 1,919 once invoked, about $0.0005 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
digest
NOTE: this skill requires connected chat / email / project-tracker / docs MCPs, which are typically only present in Cowork — the Code variant exists for parity but most users will want the Cowork variant. Cross-tool rollup of what's pending, what shipped, and what's blocked across chat / email / project tracker /…
roadmap
Render a per-project feature roadmap — a Band×Status grid (Shipped/Current/Next/Later × done/in-progress/blocked/buildable) synthesized from CLAUDE.md + PROGRESS.md, persisted to a committed ROADMAP.md with staleness-aware refresh. Modes: '/roadmap' (nearest project), '/roadmap ' (a projectslist tag), '/roadmap…
stitch
Build or verify cross-repo STITCH.md linking backend + frontends in a product group. Modes: create, verify, diff, section. Uses CODEMAPs as drift source by default. Trigger: '/stitch create ', '/stitch diff '.
aria-assist
Morning product-management review across all your projects. Use when the user runs /aria-assist, asks for a morning review / daily PM digest / "what should I do today across my products", or when the launchd job fires it headless. Two modes - generate (headless: read facts, deep-review ACTIVE projects, apply logged…
backlog
View and manage pending backlog items. Use when user says '/backlog', '/backlog insights', '/backlog clear', 'what's pending', 'show backlogs', 'check backlog status'. (Code port — ADR-094.).
parallel-feature-development
Coordinate parallel feature development with file ownership strategies, conflict avoidance rules, and integration patterns for multi-agent implementation. Use this skill when decomposing a large feature into independent work streams, when two or more agents need to implement different layers of the same system…