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/dheerg/swarms/gate-presentationnpx skills add DheerG/swarms --skill gate-presentationgit clone --depth 1 https://github.com/DheerG/swarmsWrote 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/dheerg/swarms/gate-presentation)<a href="https://agentmods.dev/skills/dheerg/swarms/gate-presentation"><img src="https://agentmods.dev/badge/skills/dheerg/swarms/gate-presentation.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 | $0.00081 | $0.01576 |
| Opus 5 | $0.00041 | $0.00788 |
| Sonnet 5 | $0.00016 | $0.00315 |
| Haiku 4.5 | $0.00008 | $0.00158 |
Grade A, and why
gate-presentation 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 4d 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 — 55 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Return the following gate constants verbatim to the team lead. Do not summarize or interpret — these are frozen SHAPE; the Gate Presentation transport contract in swarm:workflow-rules governs how they are rendered.
Gate Presentation Catalog
Setup gate — launch Step 2; generated workflow step 3.
- question: "How would you like to set up the team?" · header: "Setup"
- options: "Ultra — reliability (Recommended)" / "Auto-configure mode, team, and research. Full team on the stronger model — reliable rule-following." · "Balanced — lower cost" / "Same auto-config, but members run a cheaper model — less reliable rule-following."
- digest: none (no run values). preview: none.
Team gate — launch Step 4; generated workflow step 4.
- question: "Does this team look right? Suggested team: [for each member: name — 2–4-word role axis]." · header: "Team"
- options: "Yes, looks good" / "Proceed with this team composition" · "I want to adjust" / "Swap a member, add more members, or remove some"
- digest: the roster as handles — every member's name plus a 2–4-word axis compressed from its identity line (a compression, not the identity line itself).
- preview (same on both options): the full roster as formatted markdown — numbered, every member with its full one-line identity. The full identity lines live here only.
Plan gate — launch Step 7; generated workflow step 5; /swarm:refine Step 3.
- question: "Is this plan final, or do you have remaining inputs? [digest]" · header: "Confirm"
- options: "Launch the team" / "Plan is final — start creating the team now" · "I have changes" / launch variant: "Adjust outcomes, mode, members, tier, or research first." — refine variant: "Adjust outcomes first (roster and tier are fixed; the diff base cannot be changed from this prompt)."
- digest, launch variant (delta-forward): ship definition (one line — its first appearance); compressed outcomes line (adjustable here, so decision content); scope line only if the reflection fork kept a pin (Option A: "[their word] specifically"); then handles — name-only roster, cost tier token, mode.
- digest, refine variant: diff base first (the silently-inferred correctness pivot); PR state (URL or
(no open PR detected)); when the base is the default-branch fallback, inline:no PR detected — diff base falls back to the repo's default branch (<resolved-default>). Verify before launch.; compressed outcomes line; scope line only if the reflection fork kept a pin (Option A); branch under review; the one-line diff stat. Fixed fields (mode, tier, roster, phase arc, ship definition) are not decision content here — preview only. - preview (same on both options): elaboration only, never a restatement of the digest — the outcomes verbatim (numbered, exact confirmed wording), the full roster with identity lines, and (refine variant) the fixed fields and ship definition; plus the closing line
Rules: Active. Fields the digest carries at their only fidelity (mode and tier tokens in the launch variant; the digest one-liners) do not repeat here.
Approve gate — every mode's Approve phase.
- question: "Does this [subject] look right?" — [subject] is fixed by the mode skill: approach (code, general), diagnosis (triage), direction (writing) · header: "Approve"
- options: "Yes, proceed" / "Approve the [subject] — the team proceeds to Execute" · "I have changes" / "Adjust the [subject] before the team proceeds"
- digest: none beyond the question — the synthesis is unsplittable verbatim content; a lead-authored compression would breach the verbatim-relay rule.
- preview (same on both options): the facilitator's CONVERGED synthesis verbatim. The main-window relay still happens per the phase arc (it is the scrollback record and the AFK projection's source); the preview is what makes the synthesis readable inside the modal.
- standalone-sufficiency carve-out: this gate's sufficiency source is the verbatim relay that immediately precedes the modal, not a digest — compressing the synthesis into the question would paraphrase it.
- declared AFK carrier: every re-emission of this gate (an AFK restatement, a pulse re-emission, post-compaction) re-carries the synthesis verbatim. The lead's same-turn AFK restatement reuses the just-relayed verbatim synthesis in context; a cross-turn re-emission (the pulse's, or post-compaction) re-reads it from the facilitator's on-disk transcript (the Team Lead disk-read rule) unconditionally — drift is a cross-turn property, and silent compaction drift defeats self-detection. On that cross-turn path, never reconstruct it from memory; if the disk read fails, do not fall back to an in-context copy — surface the failure once, then hold the gate.
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.
- 4d ago First seen · 55 lines · 81 tokens per session scan A 4031ff6dd628
gate-presentation is a skill published in the GitHub repository DheerG/swarms (86 stars, last pushed 1mo ago), licensed MIT. It adds 81 tokens to every session and 1,576 once invoked, about $0.0004 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
soloflow
Grind on one goal concurrently in a Solo workspace.
setup
Use when installing or refreshing sound taste rules in a project — an explicit "sound setup", starting work in a repo with no .claude/sound/, or a taste-not-defined nudge. Also use to re-run after a stack change (new framework, new package in a monorepo).
moo-authoring
Use when adding or changing a skill, hook, or runtime file in this repo — including deciding which unit a new capability should be.
compose
Plan the skills an ask needs, invoke each as its condition arrives. Use when the ask has multiple aspects that different skills carry, or the user named multiple skills.
consult
Bring how named experts would think about the question. Use when the user wants expert thinking on a tradeoff, choice, or stuck problem — not a verdict.
explain
Put one settled thing in the user's head, plainly. Use when the user does not hold something already settled.