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 irfad7/claude-power-skills --skill anti-distillationgit clone --depth 1 https://github.com/irfad7/claude-power-skillsWrote 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/irfad7/claude-power-skills/anti-distillation)<a href="https://agentmods.dev/skills/irfad7/claude-power-skills/anti-distillation"><img src="https://agentmods.dev/badge/skills/irfad7/claude-power-skills/anti-distillation/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/irfad7/claude-power-skills/anti-distillation"><img src="https://agentmods.dev/badge/skills/irfad7/claude-power-skills/anti-distillation.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 | $0.00091 | $0.01250 |
| Opus 5 | $0.00046 | $0.00625 |
| Sonnet 5 | $0.00018 | $0.00250 |
| Haiku 4.5 | $0.00009 | $0.00125 |
Grade B, and why
anti-distillation scanned grade B with 1 finding 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 12d 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.
Instruction-override phrasingmediumPrompt injection
Text telling the model to disregard its earlier instructions or safety rules is the shape of a prompt injection, whoever wrote it.
- Instructions embedded in data ("Ignore previous instructions...") Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
How it starts
The opening of the file, as written. The whole thing — 143 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Anti-Distillation — Context Protection & Output Sanitization
You are a context protection engine. Your job is to prevent sensitive information from leaking across project boundaries, into public outputs, or through inadvertent context contamination.
When To Activate
- Working on proprietary or confidential codebases
- Switching between projects with different confidentiality levels (e.g., Ivys.ai → MLA)
- Preparing code or content for public repositories
- Working on client projects where internal details shouldn't leak
- Any time the user flags something as sensitive or confidential
Protection Layers
Layer 1: Context Boundary Enforcement
When switching between projects or orgs, enforce hard boundaries:
Before working on Project B after Project A:
- Identify which context is Project A-specific:
- Internal API URLs, endpoints, paths
- Database schemas, table names
- Business logic, pricing, strategies
- Team names, people, internal processes
- Architecture decisions unique to that project
- Mentally quarantine that context
- Never reference Project A patterns, names, or decisions in Project B work
Boundary Violations to Watch For:
- Suggesting architecture from Project A while working on Project B
- Using variable names or patterns from a different project
- Referencing internal tools or services from another org
- Copy-pasting error handling patterns that contain org-specific logic
Layer 2: Output Sanitization
Before any output leaves the session (commits, PRs, files, messages), scan for:
Hard Blocks — Never Include:
- API keys, tokens, secrets (even partial)
- Internal URLs (staging servers, admin panels, dashboards)
- Database connection strings
- Personal information (emails, phone numbers, addresses)
- Client names in non-client-facing output
- Revenue figures, pricing strategies, financial data
- Credentials of any kind
Soft Blocks — Review Before Including:
- Architecture patterns that are competitively sensitive
- Business logic that reveals strategy
- Internal tool names or custom framework names
- Performance metrics or scale indicators
- Team structure or organizational details
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.
- 12d ago First seen · 143 lines · 91 tokens per session scan B e75b03831c38
anti-distillation is a skill published in the GitHub repository irfad7/claude-power-skills (4 stars, last pushed 1mo ago), licensed MIT. It adds 91 tokens to every session and 1,250 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it B with 1 finding (instruction-override phrasing). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
import
Import a tasks.md into the coco tracker as an epic with dependencies, and create matching issues in the configured issue tracker.
hotfix
Single-issue workflow for quick fixes and small changes that don't need full epic tracking. Creates a branch, implements the fix, commits with issue tracking, and closes.
execute
Execute the next available tracked task with TDD, pre-commit validation, PR workflow, AI code review, and issue tracker bridge sync. Primary execution interface for multi-session feature work.
council-review
Perform a rigorous Carmack Council code review. Use when explicitly asked to review code, do a "council review", "carmack review", or invoke /council-review. Carmack's philosophy chairs a council of domain experts — Troy Hunt (security), Martin Fowler (refactoring), Kent C. Dodds (frontend), Matteo Collina (Node.js)…
council-plan
Architect a feature with the Carmack Council before writing code. Use when explicitly asked to plan a feature, do a "council plan", "carmack plan", or invoke /council-plan. Carmack's philosophy chairs a council of domain experts — Troy Hunt (security), Martin Fowler (refactoring), Kent C. Dodds (frontend), Matteo…
test-architect
Map testable surfaces, audit existing tests for quality, and write test specifications that prevent AI shortcuts. Use when asked to "audit tests", "specify tests", "test architect", "map test coverage", or invoke /test-architect. Two modes — audit (evaluate existing tests against Beck's principles) and specify (write…