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 softspark/ai-toolkit --skill councilgit clone --depth 1 https://github.com/softspark/ai-toolkitWrote 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/softspark/ai-toolkit/council)<a href="https://agentmods.dev/skills/softspark/ai-toolkit/council"><img src="https://agentmods.dev/badge/skills/softspark/ai-toolkit/council/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/softspark/ai-toolkit/council"><img src="https://agentmods.dev/badge/skills/softspark/ai-toolkit/council.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 | $0.00029 | $0.01215 |
| Opus 5.5 | $0.00012 | $0.00486 |
| Sonnet 5.5 | $0.00006 | $0.00243 |
| Haiku 4.5 | $0.00003 | $0.00121 |
Grade A, and why
council 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 — 134 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Decision Council
$ARGUMENTS
Evaluate a decision from 4 distinct perspectives using parallel sub-agents, then synthesize a weighted recommendation.
Usage
/council [decision question or proposal to evaluate]
What This Command Does
- Parses the decision question from user input
- Gathers codebase context relevant to the decision
- Spawns 4 parallel sub-agents — each analyzing from a different perspective
- Synthesizes all perspectives into a structured recommendation with decision matrix
MANDATORY: You MUST use the Agent tool
DO NOT analyze the decision yourself. Your job is to parse the question, spawn 4 sub-agents in parallel via the Agent tool, and synthesize their outputs. If you attempt to provide all perspectives inline, you have failed.
Process
1. Parse the Decision
Extract from the user's input:
- The proposal — what is being considered
- The alternatives — what the options are (if not stated, infer the implicit alternative: status quo)
- The context — what project/codebase/situation this applies to
If the question is too vague to evaluate, ask ONE clarifying question before proceeding.
2. Gather Context
Before spawning agents, quickly scan the codebase for relevant context:
- Current tech stack, dependencies, patterns in use
- Existing code related to the decision area
- Configuration, infrastructure, or architectural choices already made
Include this context in each sub-agent prompt so perspectives are grounded in reality.
3. Spawn 4 Sub-Agents in Parallel (REQUIRED)
Call the Agent tool 4 times in a single response — all agents MUST launch in the same message to run in parallel.
| Agent | Role | Directive |
|---|---|---|
| Advocate | Case FOR | Present the strongest case FOR the proposal. Find evidence, benefits, success stories. Reference relevant code/patterns in the codebase. Be persuasive but honest. |
| Critic | Case AGAINST | Present the strongest case AGAINST. Find risks, hidden costs, failure modes, alternatives that might be better. Be thorough but fair. |
| Pragmatist | Trade-offs | Evaluate practical trade-offs: implementation cost, timeline, team capacity, maintenance burden, migration risk, operational complexity. Ground estimates in the actual codebase size and patterns. |
| User-Proxy | User impact | Consider end-user and customer impact: UX changes, downtime during migration, performance implications, breaking changes, adoption friction, documentation needs. |
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 First seen · 134 lines · 29 tokens per session scan A ecb8dc194280
council is a skill published in the GitHub repository softspark/ai-toolkit (180 stars, last pushed yesterday), licensed Apache-2.0. It adds 29 tokens to every session and 1,215 once invoked, about $0.0001 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 2026-09-03.
Other skills, from other repositories
cline-fix-volatile-msg
Ladder-aware Cline Anthropic caching — verify the rolling read/write ladder on the wire, then add the tools breakpoint and tune TTL. Updated for the 2026-08 AI-SDK monorepo.
continue-gemini-explicit
Continue's Gemini provider doesn't use the cachedContents API at all. Add explicit caching for sessions over the minimum token threshold.
roo-fix-volatile-msg
Ladder-aware Roo Code Anthropic caching — verify the rolling read/write ladder on the wire, then close the real gaps (Vertex 4-block budget, MiniMax path).
aider-1h-ttl
Aider uses 5min TTL by default and works around long pauses with keepalive pings. Wire up the 1h TTL beta instead.
cline-openai-cache-key
Cline OpenAI native provider sends no promptcachekey. Add a stable per-task key so cachedtokens stops being zero.
continue-enable-defaults
Continue's prompt caching is opt-in via config and off by default. Flip the default to systemAndTools.