| name | lean-report-writer |
| description | Turn sustainability audit findings into a polished written report (Word/PDF/markdown) — executive narrative, findings detail, remediation roadmap, with optional ESG/CSRD-flavored framing on request. Use this skill whenever the user wants a written sustainability/efficiency report, documentation of audit results, content for an ESG or engineering report, or a lean-router pipeline ends in a document. Consumes one or more lean-findings.json files from any Lean Agentic AI Skills audit; never audits anything itself. |
Green Report Writer
Pure consumer skill. Input: one or more lean-findings.json files (merge mechanically if multiple: concatenate findings, union passed/out_of_scope, keep all methodology paragraphs). Output: a report document.
Prerequisite (platform-aware): default to markdown unless the user asks for a formal document. For .docx/.pdf output, use your environment's document skills where present (e.g. Claude's docx/pdf skills under /mnt/skills/public/); on other agent platforms use available document tooling (python-docx, reportlab, or equivalent).
Report structure
Write in prose — findings data becomes narrative, not pasted JSON:
- Executive summary — scope in one sentence, findings counts by severity, the single biggest lever, one honest sentence on what's already healthy.
- Scope & methodology — mandatory, verbatim-faithful to every
methodology paragraph and scope_notes. A report without its methodology is greenwashing; this section cannot be cut even if the user asks for "just the findings" (offer a short version that still carries a one-line methodology note).
- Findings — high severity as full subsections (issue → evidence → impact → remediation, with pattern citations as links); medium grouped by theme; low as a summary paragraph. Translate SCI terms inline for lay readers.
- Remediation roadmap — sequence by effort-vs-severity: quick wins (low effort, ≥medium severity) first with a "week one" framing, then structural items.
- What's already working — the
passed list as prose; genuine credit, not filler.
- Limitations & next steps —
out_of_scope items reframed as the follow-up plan, plus pointers to real measurement tooling (CodeCarbon, Cloud Carbon Footprint, Kepler, Impact Framework) for anything the audit could only assess directionally.
ESG/CSRD framing (only when requested)
Findings can inform ESRS E1-style narrative (actions taken, planned resource allocation) but this skill does NOT produce compliance-grade disclosures: no emissions inventories, no Scope 1/2/3 figures, no assurance-ready claims. State this limit in the document itself and recommend the user's sustainability/audit function own final disclosure text.
Numbers discipline
Only figures present in evidence appear in the report. No derived percentages, no CO2 grams, no cost extrapolations beyond what a findings file contains. Where a section feels thin without numbers, the honest sentence is "directional finding — measurement recommended (see Limitations)".
Cost signal rendering
Cost signal integration. Where findings carry cost_signal, weave the counted driver into each finding's paragraph (e.g. "...the driver here is bytes egress per visit, currently ~4.2 GB per page load..."). Never estimate spend without a supplied billing rate. Reports MUST preserve the three-axis honesty split: energy and carbon directional, cost driver counted.
