summarizing-evidence

Generate strict-schema JSON summaries of Fianu findings, violations, and attestations.

1|Updated Jun 11, 2026
One-click install
npx skills add https://github.com/fianulabs/fianu-skills --skill summarizing-evidence-fianulabs
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: summarizing-evidence
Source: https://github.com/fianulabs/fianu-skills/tree/main/skills/summarizing-evidence
Command: npx skills add https://github.com/fianulabs/fianu-skills --skill summarizing-evidence-fianulabs

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Security and compliance reviewers need to grasp the meaning of a Fianu finding, violation, or attestation quickly inside the evidence panel, without reading raw evidence payloads or wading through verbose prose. ## Core Features & Use Cases - Strict JSON output contract: Produces a single JSON object with a 2-4 sentence plain-prose summary and an ordered list of up to 5 actionable next steps, with no markdown fences or commentary. - Subject-specific framing: Tailors the summary content to the evidence type — findings (location, severity, rule id), violations (broken policy, flagged control), and attestations (control result, asset, commit). - Grounded, non-editorialized content: Cites only identifiers present in the context (rule ids, file paths, commit SHAs, CVE/CWE numbers) and never invents details or adds subjective judgment. - Use Case: A developer opens a policy violation in the Fianu console and immediately sees a concise summary stating which policy was broken and at what severity, plus a prioritized next step such as upgrading a vulnerable dependency. ## Quick Start Summarize this Fianu finding as JSON for the evidence panel, including prioritized next steps grounded in the provided context.

Frequently Asked Questions about summarizing-evidence

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I summarize a Fianu finding or violation as JSON?▼

Provide the finding, violation, or attestation context and request a summary. The output is a single JSON object with a summary field of 2-4 plain-prose sentences and a nextSteps array of up to five prioritized actions, with no markdown fences or extra commentary.

What JSON schema does the evidence summary output follow?▼

The schema has a required summary string and a nextSteps array of 0-5 entries. Each step contains an imperative action one-liner plus optional details and priority fields, where priority is high, medium, or low. Empty fields are omitted rather than emitted as empty strings.

Can the summary include identifiers not present in the evidence context?▼

No. The summary only cites identifiers that already exist in the context, such as rule ids, file paths, short commit SHAs, and CVE or CWE numbers. If a field is missing, it is left out rather than guessed, and thin context is stated plainly.

How are next steps prioritized in the evidence summary?▼

Priority reflects blast radius and urgency: high for actively exploitable or production-blocking issues, medium for clear non-urgent remediation, and low for hygiene. Steps are ordered by importance, and vague items like investigate further are omitted entirely.

When should I not use this evidence summarization approach?▼

Do not use it when you need to fetch or regenerate the summary through the API, since this skill only defines the output contract. For the endpoint call itself, use the working-with-attestations skill covering the summarize endpoint.