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 jxoesneon/Ciel --skill context-budgetgit clone --depth 1 https://github.com/jxoesneon/CielWrote 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/jxoesneon/ciel/context-budget)<a href="https://agentmods.dev/skills/jxoesneon/ciel/context-budget"><img src="https://agentmods.dev/badge/skills/jxoesneon/ciel/context-budget/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/jxoesneon/ciel/context-budget"><img src="https://agentmods.dev/badge/skills/jxoesneon/ciel/context-budget.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 | $0.00022 | $0.00539 |
| Opus 5 | $0.00011 | $0.00269 |
| Sonnet 5 | $0.00004 | $0.00108 |
| Haiku 4.5 | $0.00002 | $0.00054 |
Grade A, and why
context-budget 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 8d 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.
What it actually says
CIEL ADAPTATION: Context Budget (Context Audit)
This skill provides the Orchestrator with a proactive mechanism to monitor and optimize the "Permanent Context" overhead of the system. It identifies bloat in agent definitions, redundant skills, and over-subscribed MCP servers.
Audit Workflow
Phase 1: Resource Inventory
The Orchestrator estimates the token footprint of:
- Agents: Line counts and description frontmatter.
- Skills: Total tokens across
SKILL.mdfiles. - Rules: Global and project-level rule files.
- MCP Servers: Estimated schema overhead (~500 tokens per tool).
- Project State:
CLAUDE.mdandTodoWritelength.
Phase 2: Classification & Triage
Components are sorted into three categories:
- Essential: Backs an active command or matches project type.
- On-Demand: Lazy-loadable or domain-specific.
- Redundant: Overlapping logic or rarely used.
Phase 3: Optimization Recommendations
The Auditor produces a prioritized list of token-saving actions:
- Agent Thinning: Reducing verbose descriptions.
- MCP Pruning: Removing servers that wrap simple CLI commands.
- Rule Consolidation: Merging overlapping language rules.
The 40% Rule
The Orchestrator should aim to keep the "System Overhead" (permanent context) below 40% of the model's total context window to ensure sufficient headroom for conversation and tool results.
Anti-Patterns
- Invisible Bloat: Keeping 20+ MCP servers active "just in case."
- Verbose Frontmatter: Agent descriptions that read like long tutorials.
- Rule Duplication: Repeating global rules in project-specific files.
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.
- 8d ago First seen · 64 lines · 22 tokens per session scan A 8eda01c13f24
context-budget is a skill published in the GitHub repository jxoesneon/Ciel (1 stars, last pushed 4d ago), licensed Apache-2.0. It adds 22 tokens to every session and 539 once invoked, about $0.0001 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 2026-09-03.
Other skills, from other repositories
meeting-notes
OwnPilot official general skill for converting rough meeting notes, transcripts, and call summaries into decisions, action items, owners, risks, and follow-ups. Use when the user provides meeting content or asks for notes.
next-action
Use when the user wants a next-action recommendation based on current state — reads handoff/git/lessons/STATE and proposes top-3 by impact. Trigger: '/next', 'what should I do next', 'next action'. Proposes only, never executes.
agenticx-automation-crontask
Build and maintain Near Desktop scheduled (cron) tasks — default workspace /.agenticx/crontask, scheduletask tool, execution contract, and user-facing output. Use when the user wants recurring automation, crontab-style jobs, or to author/fix automation task prompts.
day-close
A day-ending protocol that reviews and records the required closing steps for the workday in IWE.
code-dev-workflow
Near codedev harness — Explore, Read, Author phases with outline-first context discipline.
lov-canonical-store
A central file and command-line workflow for maintaining one deduplicated vocabulary list as the source for other apps.