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 ychampion/cskill-agents --skill capability-aware-offload-skippinggit clone --depth 1 https://github.com/ychampion/cskill-agentsWrote 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/ychampion/cskill-agents/capability-aware-offload-skipping)<a href="https://agentmods.dev/skills/ychampion/cskill-agents/capability-aware-offload-skipping"><img src="https://agentmods.dev/badge/skills/ychampion/cskill-agents/capability-aware-offload-skipping/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/ychampion/cskill-agents/capability-aware-offload-skipping"><img src="https://agentmods.dev/badge/skills/ychampion/cskill-agents/capability-aware-offload-skipping.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 | $0.00028 | $0.00388 |
| Opus 5 | $0.00014 | $0.00194 |
| Sonnet 5 | $0.00006 | $0.00078 |
| Haiku 4.5 | $0.00003 | $0.00039 |
Grade A, and why
capability-aware-offload-skipping 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 9d 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
SKILL: Capability-Aware Offload Skipping
Domain: tool-result-storage Trigger: Use when a generic result-offload layer wraps many tools, but some tools already enforce their own output bounds or would recurse if wrapped again. Source Pattern: Distilled from reviewed output-budget enforcement and selective offload-policy implementations.
Core Method
Before applying a generic persistence or preview wrapper, inspect each tool's declared behavior and opt out the tools whose own contract already bounds output or would create a circular readback loop. Treat the skip as a stable policy decision rather than a one-off exception so identical tools keep the same behavior across turns. This preserves the generic offload path where it adds value while avoiding redundant or self-referential wrapping.
Key Rules
- Check declared limits and self-bounded reader behavior before applying a generic offload layer.
- Keep the skip decision stable across turns so identical tool results do not oscillate between wrapped and unwrapped behavior.
- Mark skipped results consistently when broader budgeting logic depends on deterministic treatment across turns.
- Prefer tool-specific contracts over wrapper-level force when the wrapper would just reintroduce the same content through another tool.
Example Application
If a read-style tool already caps output aggressively, exclude it from a generic file-offload wrapper; otherwise the wrapper would persist bounded content and ask the model to read it back through the same tool, adding complexity with no real gain.
Anti-Patterns (What NOT to do)
- Do not wrap every tool uniformly just because a generic offload helper exists; some tools already solve the problem locally.
- Do not let skip decisions fluctuate between turns, or identical tool results will produce inconsistent prompt prefixes and cache behavior.
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.
- 9d ago First seen · 28 lines · 28 tokens per session scan A c68d8754db74
capability-aware-offload-skipping is a skill published in the GitHub repository ychampion/cskill-agents (36 stars, last pushed 5mo ago), licensed MIT. It adds 28 tokens to every session and 388 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 2026-08-30.
Other skills, from other repositories
grape
Use Grape MCP for Codex context continuity in coding repositories. Use when a task needs repeated-turn context, omitted context restore, stale-context checks, invalidation checks, or safe continuity across branch and dirty-worktree changes.
opensrc
Fetch dependency source code to give AI agents deeper implementation context. Use when the agent needs to understand how a library works internally, read source code for a package, fetch implementation details for a dependency, or explore how an npm/PyPI/crates.io package is built. Triggers include "fetch source for"…
review
Validate plans, execution, or PRs against wish criteria — returns SHIP / FIX-FIRST / BLOCKED with severity-tagged gaps.
work
Execute an approved wish plan — orchestrate subagents per task group with fix loops, validation, and review handoff.
brainstorm
Explore ambiguous or early-stage ideas interactively — tracks wish-readiness and crystallizes into a design for wish.
genie
Entry point for Genie operations — routes bug reports, questions, and operational commands, resumes existing lifecycle state, and orchestrates work that needs durable planning or coordination. Other ordinary requests bypass the lifecycle with a one-line notice unless the user asks for Genie.