validate-output — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited validate-output (Agent Skill) and scored it 96/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 1 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 1 flagged
The text {match} tells the agent to skip the normal "ask the user first" gate. Used adversarially it removes the human-in-the-loop check before destructive or sensitive actions, turning a normally-gated agent into a fire-and-forget executor.
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Validate marketing content against expected structural schemas to ensure completeness, formatting consistency, and production-readiness. Checks required sections, word count ranges, markdown formatting compliance, placeholder text detection (unfilled template variables, lorem ipsum, TBD markers), and content-CTA consistency. Supports eight built-in schemas for common marketing content types plus custom schemas for brand-specific templates.
This command catches the structural and formatting issues that quality evaluation misses — the missing H2 that breaks SEO, the placeholder "[INSERT COMPANY NAME]" that slipped through, the blog post that is 300 words short of the brief requirement, or the email that has a CTA promising a demo but the body talks about a whitepaper. It is designed to be run as a final pre-publication check after content quality has been evaluated via /digital-marketing-pro:eval-content.
The user must provide (or will be prompted for):
blog_post, email, ad_copy, social_post, landing_page, press_release, content_brief, campaign_plan — or a file path to a custom JSON schema. If omitted, the validator auto-detects the most likely schema based on content structure, length, and formatting patterns. Custom schemas follow the format defined in skills/context-engine/eval-framework-guide.md~/.claude-marketing/brands/_active-brand.json for the active slug, then load ~/.claude-marketing/brands/{slug}/profile.json. Apply brand formatting standards and content requirements. Also check for guidelines at ~/.claude-marketing/brands/{slug}/guidelines/_manifest.json — if present, load template definitions from templates/ that may define brand-specific required sections, word count ranges, and formatting rules. Check for custom schemas at ~/.claude-marketing/brands/{slug}/schemas/. Check for agency SOPs at ~/.claude-marketing/sops/. If no brand exists, ask: "Set up a brand first (/digital-marketing-pro:brand-setup)?" — or proceed with defaults.scripts/output-validator.py --action list-schemas to get all available schemas, then select the most appropriate one based on content characteristics (length, structure, formatting patterns). Report which schema was selected and why, so the user can override if the selection was wrong.scripts/output-validator.py --action validate --text "{content}" --schema {builtin_schema_name} for one of the eight built-in schemas, or --custom-schema {path/to/schema.json} for a custom schema file (the two flags are distinct — --schema takes a built-in name only, --custom-schema takes a file path). The validator checks:{placeholder}, [PLACEHOLDER], [INSERT X], TODO, TBD, FIXME, Lorem ipsum, xxx, ACME Corp used as placeholder), partial completions, and obviously templated content that was not customizedname, sections (array of section definitions with name, required flag, min/max word count), total_word_count (min/max), formatting_rules, placeholder_patternsA structured validation report containing:
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.