AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified MIT Self-run

Generate

skill-melodic-software-claude-code-plugins-generate · by melodic-software

Build a source-backed AI industry briefing from official vendor publications, configured RSS feeds, GitHub releases, reputable secondary reporting, and user-supplied URLs. Use when: 'ai briefing', 'ai news', 'what's new in AI', 'catch me up on AI', 'prep for AI meeting', 'AI roundup', or 'generate AI slides'.

No reviews yet
0 installs
28 views
0.0% view→install

Install

$ agentstack add skill-melodic-software-claude-code-plugins-generate

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No 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 Used
  • 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.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-melodic-software-claude-code-plugins-generate)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
1mo ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

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 →
Are you the author of Generate? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Variables

Arguments: $ARGUMENTS Configured active profile: ${user_config.active_profile}

Purpose

Collect, deduplicate, rank, and present current AI-industry developments without depending on organization-specific configuration. Markdown is the default output. HTML and PPTX are explicit, optional outputs backed by the bundled deterministic build pipeline.

Source and access policy

Use only interfaces that authorize automated access:

  1. Official vendor blogs, documentation, changelogs, and release notes.
  2. GitHub release pages and repository APIs.
  3. RSS/Atom feeds listed by the selected profile.
  4. Reputable secondary reporting, corroborated by a primary source when practical.
  5. User-supplied public URLs and context.

Automated X/Twitter collection is disabled. Do not crawl, scrape, scroll, navigate, fetch, or extract X pages through browser automation, DOM scripts, unofficial endpoints, or Grok. Do not refresh following graphs or profile metadata. X's current terms expressly prohibit crawling or scraping without prior written consent: .

If the user supplies an X URL, preserve it as user-provided citation metadata but do not programmatically retrieve it. Ask for the relevant text when it is necessary, and seek a non-X primary source for corroboration. This plugin does not configure a paid X API.

Playwright is allowed only for deterministic rendering and inspection of locally generated HTML/PDF artifacts. It is not a collection provider.

Arguments

| Flag | Default | Effect | |---|---|---| | --profile | Configured active_profile, then default | Select a named profile for this invocation. | | --since | 14d | Accept 1d, 3d, 7d, 14d, 30d, or an ISO date. | | --providers | All | Limit collection to comma-separated provider buckets. | | --extras / --no-extras | On | Include or omit robotics, science, and novel applications. | | --format | markdown | Emit markdown, HTML, or PPTX. Non-markdown formats require prior /ai-briefing:setup apply install-build-deps. | | --yes / -y | Off | Skip the pre-execution confirmation gate. Required for headless runs. |

Actions:

  • Default: collect, deduplicate, rank, cite, and emit a briefing.
  • retro --meeting : record acted/noted/skipped feedback for an archived briefing.
  • search "": search archived briefing markdown and return cited matches.

Profile resolution

Files at .claude/ai-briefing/ form the default profile. Each .claude/ai-briefing// directory is a named profile overlay. Resolve the active profile in this order:

  1. --profile for this invocation.
  2. The rendered ${user_config.active_profile} value when it is non-empty.
  3. default.

Require a 1-63 character lowercase-kebab profile name and reject reserved Windows device names before constructing any profile path.

Resolve that value in the skill before launching tools. Explicitly pass the result to every build subprocess; do not assume plugin configuration is inherited by an arbitrary shell:

AI_BRIEFING_PROFILE="$PROFILE" node "${CLAUDE_PLUGIN_DATA}/runtime/build/run.js"

Never ask the consumer to export the variable globally.

A profile may contain:

| Artifact | Purpose | |---|---| | sources.md | Approved RSS/Atom feeds, official release pages, GitHub repositories, and user-supplied URLs. | | audience.md | Optional stack/audience lens for impact annotations. | | brand.json | Optional declarative deck brand overlay. |

Default run

  1. Parse and validate arguments. Reject unknown formats and malformed dates before any

collection. Resolve the selected profile and read sources.md when present.

  1. Unless --yes is present, show the timeframe, provider scope, source classes, selected

profile, and requested output format. Ask for confirmation before outbound collection.

  1. Search official vendor publications and GitHub releases first. Read configured feeds and

permitted user-supplied non-X URLs. Use reputable secondary reporting to fill gaps and corroborate claims. Record source URL, publisher, publication date, and retrieval date.

  1. For every item, require at least one working source URL. Prefer the primary announcement;

label claims that remain secondary-only. Exclude content outside the requested window.

  1. Deduplicate by canonical URL and normalized event identity. Merge corroborating sources

into one item rather than repeating the same announcement.

  1. Categorize into provider buckets and rank by practical impact, availability, novelty,

source confidence, and relevance to audience.md. Never invent details or silently turn rumors into facts.

  1. Write the current markdown briefing and update the seen-item registry under the selected

profile's ${CLAUDE_PLUGIN_DATA} state directory. Keep tracked profile configuration in the project; never write curated configuration into plugin data.

  1. For --format html or --format slides, require the optional build tree installed by

/ai-briefing:setup apply install-build-deps. Run the staged build pipeline against the emitted markdown with AI_BRIEFING_PROFILE="$PROFILE" set on the launched process. Playwright may open only generated local HTML for PDF rendering and layout validation. Surface missing prerequisites with the exact setup command; do not silently fall back to another browser provider.

  1. Report collected, excluded, deduplicated, and secondary-only counts plus the output path.

Make warnings visible, including unreachable sources and missing optional build tooling.

Output contract

The markdown briefing contains:

  • Run metadata: profile, window, retrieval time, and source classes used.
  • Provider sections with HIGH, MEDIUM, and LOW tiers.
  • Item title, concise evidence-backed summary, practical impact, publication date, and

source links.

  • A visible caveats section for inaccessible sources, secondary-only claims, and gaps.
  • An optional EXTRAS section when enabled.

Machine-local state and generated artifacts live under ${CLAUDE_PLUGIN_DATA}//. Tracked source, audience, and brand configuration lives under .claude/ai-briefing/[/] in the consumer repository.

Quality gates

  • Every factual item has a source URL and date.
  • Primary sources are preferred and clearly distinguished from secondary reporting.
  • No automated X access occurs, including through browser tools or indirect DOM scripts.
  • The same event appears once, with corroborating links attached.
  • Missing dependencies and partial collection are visible; no black-box degradation.
  • HTML/PDF browser use is local-render-only.

References

  • references/audience-defaults.md — default ranking lens and profile overlay.
  • references/build-pipeline.md — deterministic HTML/PDF/PPTX generation and validation.
  • references/slide-generation.md — slide structure and optional build prerequisites.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.