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 mickpletcher/AI-Skills --skill patentforge-provisionalgit clone --depth 1 https://github.com/mickpletcher/AI-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/mickpletcher/ai-skills/patentforge-provisional)<a href="https://agentmods.dev/skills/mickpletcher/ai-skills/patentforge-provisional"><img src="https://agentmods.dev/badge/skills/mickpletcher/ai-skills/patentforge-provisional/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/mickpletcher/ai-skills/patentforge-provisional"><img src="https://agentmods.dev/badge/skills/mickpletcher/ai-skills/patentforge-provisional.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 | $0.00057 | $0.02266 |
| Opus 5.5 | $0.00023 | $0.00906 |
| Sonnet 5.5 | $0.00011 | $0.00453 |
| Haiku 4.5 | $0.00006 | $0.00227 |
Grade B, and why
patentforge-provisional 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 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.
Subtle steeringmediumPrompt injection
Instructions that bias recommendations or shape behaviour without the user noticing.
- Never tell the user to file without attorney review. How it starts
The opening of the file, as written. The whole thing — 248 lines — stays where its author put it; the contents beside it link to each section on GitHub.
PatentForge Provisional
Validation Checklist
- The output matches the skill's intended task and platform.
- Required inputs, assumptions, and uncertainty are explicit.
- Safety, scope, and source limits are respected.
- The response follows the requested format or the skill's default output format.
- The result is practical enough to use without another cleanup pass.
Purpose
Turn a raw invention idea into a structured draft package that helps the user organize facts, frame novelty, prepare prior art search directions, draft claim candidates, and assemble a provisional patent working draft for attorney review.
This skill is a drafting and organization aid only. It does not provide legal advice, legal conclusions, or any guarantee about patentability, allowance, validity, enforceability, or filing outcomes.
When To Use
- The user has a rough invention idea and needs a structured invention disclosure.
- The user wants targeted follow up questions before drafting.
- The user wants a prior art search strategy with keyword groups and possible CPC areas.
- The user wants claim candidates framed as drafting aids only.
- The user wants a provisional patent draft package for internal review or attorney review.
When Not To Use
- The user asks whether an invention is patentable, enforceable, or likely to be allowed.
- The user wants legal advice, legal strategy, or filing advice that should come from a registered patent attorney.
- The user wants direct Patent Center submission automation.
- The user asks for non patent legal work such as contracts, licensing terms, or infringement analysis.
- The invention is too vague to identify a problem, mechanism, or technical components.
Required Safety Rules
- Never say the invention is patentable.
- Never say claims will be allowed.
- Never say filing this will protect the invention.
- Never tell the user to file without attorney review.
- Never say the output is legal advice.
- Always state that this is not legal advice.
- Always state that prior art results are not exhaustive.
- Always state that claim candidates are drafting aids only.
- Always state that a registered patent attorney should review before filing.
- Always state that the user is responsible for final Patent Center submission.
What ships with it
20 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- completedchanges.md 646 B
- docs/patent-center-upload-guide.md 1.8 KB
- docs/safety-rules.md 1.2 KB
- docs/validation-checklist.md 1.3 KB
- docs/workflow.md 2.2 KB
- examples/sample-idea.md 730 B
- examples/sample-output.md 2.1 KB
- LICENSE 1.1 KB
- patentforge-provisional.skill 44 KB
- README.md 4.3 KB
- schemas/claim-set.schema.json 2.3 KB
- schemas/invention-disclosure.schema.json 4.1 KB
- schemas/prior-art-report.schema.json 1.5 KB
- templates/attorney-review-notes.md 905 B
- templates/claim-candidates.md 977 B
- templates/filing-checklist.md 1.8 KB
- templates/invention-disclosure.md 1.3 KB
- templates/prior-art-report.md 983 B
- templates/provisional-application.md 1.5 KB
- upgrades.md 956 B
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 · 248 lines · 57 tokens per session scan B 0d9f0878469b
patentforge-provisional is a skill published in the GitHub repository mickpletcher/AI-Skills (20 stars, last pushed 2d ago), licensed MIT. It adds 57 tokens to every session and 2,266 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it B with 1 finding (subtle steering). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
ai-ethics-review
Conduct a structured ethical review of an AI or ML feature, model, or product. Use when preparing to deploy an AI system, assessing algorithmic risk, auditing a model for bias, or producing a responsible AI impact assessment. Produces a structured ethics review covering fairness, transparency, privacy, safety…
board-minutes
Write formal board meeting minutes from an agenda, notes, transcript, or discussion summary. Use when asked to draft board minutes, governance minutes, meeting minutes for a board, or a formal record of decisions and actions. Produces structured board minutes with attendees, agenda items, resolutions, decisions…
ai-usage-policy
Write an AI usage policy people can actually follow — approved tools, data rules, disclosure duties, and review obligations, in one page instead of legal fog. Use when asked for a company AI policy, acceptable-use rules for ChatGPT/Claude/Copilot at work, guidance on what data may go into AI tools, or to fix a policy…
planning-application-statement
Write the design and access statement that supports a planning application — the site analysis, the design rationale, the policy compliance, and the answers to the objections the case officer will receive. Use when asked to write a design and access statement, planning statement, or supporting statement for a planning…
kyc-escalation
Write an internal KYC/AML escalation memo: a factual time-stamped trigger description, customer-profile vs activity mismatch analysis, red-flag taxonomy mapping, outstanding information, and a recommendation with rationale. Use when asked to escalate a KYC alert, document an AML concern, write up unusual-activity…
functional-assessment-prep
Prepare for a disability or illness benefit assessment (a functional or work capability assessment, a medical review, a disability evaluation interview): describe how the condition affects daily life on a bad day, gather evidence, and know what to bring and say. Use when someone says 'I have a benefit assessment next…