Install
$ agentstack add skill-rohasnagpal-legal-ai-skills-product-liability-analyst ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Product Liability Analyst
What this does
Assesses product liability exposure on a given set of facts: what type of defect is alleged, whether the facts actually connect that defect to the harm claimed, which party or parties in the supply chain are potentially exposed, what defences might realistically be available, and the realistic exposure graded by severity. It does not draft a complaint or reply, and it does not quantify compensation beyond describing the categories of exposure.
Before you start
The facts. The product, the alleged defect, the harm or loss caused, and to whom — supplied by the user.
Which side's exposure is being assessed, or whether the question is instead whether a claimant has a viable claim. This skill is framed by default around the supply-side party's exposure, since that is what "exposure" ordinarily means, but confirm the orientation rather than assume it.
Governing law. Product liability regimes differ substantially — strict liability statutes, negligence-based regimes, a dedicated product liability chapter such as the one under India's Consumer Protection Act 2019, or a directive-based framework. Do not assume any particular regime applies. Ask, or if research tools are available and the user wants the precise legal test, cite only current, retrieved, or supplied sources; otherwise work from the general structure of the relevant defect category and flag the precise test as a verification point.
Not blocking, ask once and proceed on a reasonable default without it: the client's position in the supply chain — manufacturer, component-part maker, assembler, seller, distributor, or importer. Liability allocation across the supply chain is often regime-specific, and knowing the client's actual position sharpens which allocation questions matter.
Method
1. Classify the type of defect alleged — manufacturing defect (this unit deviated from its own design or specification), design defect (the design itself is unreasonably unsafe even when made correctly), or warning or instruction defect (inadequate warning or instructions) — in one line. Facts can implicate more than one category; say so if they do.
2. Map the facts against the elements of the applicable defect category, asserting the precise legal test only if it is sourced this session or supplied by the user. Otherwise state the general structure of the relevant category as a framework, explicitly flagged for verification against the specific law that applies.
3. Identify every party in the supply chain potentially exposed — manufacturer, component-part maker, assembler, seller, distributor, service provider — and note that how liability allocates between them (for instance, whether a seller can be liable for a manufacturing defect it did not cause, and under what conditions) is regime-specific. Flag this as a verification point rather than asserting an allocation rule from memory.
4. Assess causation on the facts actually given. Check whether the alleged defect genuinely connects to the harm claimed, or whether there is a gap or an alternative explanation — misuse, unauthorised modification, expiry, a third party's intervening act. Do not assume causation is established merely because a defect is alleged and a harm occurred.
5. Note defences potentially available on the facts — misuse, unauthorised modification, an adequate warning that was given and ignored, a state-of-the-art defence where recognised — without asserting which regime actually recognises which defence. Flag each as a verification point.
6. State the realistic exposure — the categories of relief or compensation that could be sought if liability were established (repair, replacement, refund, consequential loss, personal injury damages where applicable, a regulatory consequence), tied to compensation-quantifier's output where figures are available. Do not assert a specific figure that has not been supplied or quantified.
7. Grade the exposure using the same three-tier approach used elsewhere in this practice pack: Critical for exposure involving personal injury or safety with weak available defences, Material for a real but contestable claim, Minor for a technical or de minimis issue.
Output
1. Header. Side analysed, product, facts summarised, governing law (as stated or confirmed), date.
2. Defect classification. The type or types alleged, with the applicable framework stated and flagged for verification if not sourced this session.
3. Element-by-element analysis. A table: Element | Facts bearing on it | Satisfied / not satisfied / unclear.
4. Supply-chain exposure map. Which parties are potentially exposed, with the allocation question between them flagged as regime-dependent.
5. Causation assessment. Whether the facts actually connect the alleged defect to the harm, and any gap identified.
6. Defences potentially available. Listed, each flagged as regime-dependent.
7. Exposure and grading. The realistic exposure categories, with figures only where supplied or quantified elsewhere, graded Critical, Material, or Minor.
8. Points requiring verification. The precise legal test for the defect category, supply-chain liability allocation rules, and any defence's actual recognition under the applicable law.
Do not
Do not assert the precise legal test for a manufacturing, design, or warning defect from memory. Flag it as a general framework requiring verification unless sourced or supplied.
Do not assume a specific product liability regime applies, including India's Consumer Protection Act product liability chapter, without confirming the governing law.
Do not assume how liability allocates across the supply chain. Flag it as regime-dependent.
Do not assume causation is established from the mere fact of an alleged defect and a harm. Check that the facts actually connect them.
Do not invent an exposure figure that has not been supplied or quantified by compensation-quantifier.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: rohasnagpal
- Source: rohasnagpal/legal-ai-skills
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.