Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/indranilbanerjee/contentforgenpx agentmods add skills/indranilbanerjee/contentforge/cf-cowork-setupWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/indranilbanerjee/contentforge/cf-cowork-setup)<a href="https://agentmods.dev/skills/indranilbanerjee/contentforge/cf-cowork-setup"><img src="https://agentmods.dev/badge/skills/indranilbanerjee/contentforge/cf-cowork-setup.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Rogue Agent · line 63 Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.Fix: Remove any persistence mechanisms (cron jobs, startup scripts, state files). Skills should not maintain state across sessions without explicit user consent.
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00177 | $0.02233 |
| Opus 5 | $0.00088 | $0.01117 |
| Sonnet 5 | $0.00035 | $0.00447 |
| Haiku 4.5 | $0.00018 | $0.00223 |
Grade A, and why
cf-cowork-setup scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
Copies of this mod
1 near-identical copy found in the catalogue:
- cf-cowork-setup — 100% identical, 0 lines differ
How it starts
The opening of the file, as written. The whole thing — 163 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/contentforge:cf-cowork-setup
The one-time setup that makes ContentForge usable in Cowork by a team. Wires up the Cowork → Drive routing so generated .docx files, brand profiles, and run records actually persist somewhere the team can reach.
Why this skill exists
Cowork is the most user-friendly Anthropic surface for non-CLI users — marketers, content teams, agency staff. The natural team workflow is "everyone uses Cowork; outputs live in our shared Google Drive". But ContentForge's filesystem layer was originally designed for local Claude Code (writes to ~/Documents/ContentForge/ on the host). In Cowork that path is the Linux sandbox — gone at session end, invisible to the team.
v3.12.9 fixed this with environment-aware routing: when Cowork is detected AND a Drive MCP is configured, the output-manager agent uploads to Drive instead. This skill is the one-shot setup that ensures both conditions are true before you start producing real content.
Behavior
Step 1 — Verify Cowork environment
python scripts/plugin-metadata.py --section environment
Parse the JSON. Four branches:
environment == "cowork-sandbox" — Proceed to Step 2.
environment == "linux-sandbox-uncertain" — Detection found only weak signals (e.g., a container marker plus a Docker-style username), not a confirmed Cowork session marker. Tell the user: "Sandbox detection was uncertain — proceeding with Cowork setup to be safe. If this isn't actually Cowork, the Drive-MCP check in the next step will make that clear (no Drive MCP will be found) and you can skip the rest." Then proceed to Step 2 exactly as the cowork-sandbox branch does.
environment == "claude-code-windows" / "-mac" / "-linux" — Tell the user:
"You're running in local Claude Code, not Cowork. The Cowork-specific Drive routing isn't needed here — your files will land in
~/Documents/ContentForge/<brand>/on your host as designed. If you ALSO want Drive backups for team sharing, run/contentforge:brand-setupand pick the Google Sheets + Drive backend."
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 8d ago First seen · 163 lines · 177 tokens per session scan A c6ead918d4e5
cf-cowork-setup is a skill published in the GitHub repository indranilbanerjee/contentforge (26 stars, last pushed 22d ago), licensed MIT. It adds 177 tokens to every session and 2,233 once invoked, about $0.0009 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
continuous-improvement-loop
Run Part 12 of the engagement methodology — the continuous improvement loop that aggregates quarterly-review, customer-feedback, competitive, and operating signals into a Quarterly Product & Offering Improvement Brief for business leadership, plus fast 1-3 page ad-hoc briefs when a significant signal lands…
four-core-documents
Produce Part 3 of the 12-Part engagement: the four strategic-spine documents across 61 steps — 3.1 Business & SBU Analysis, 3.2 Segmentation Framework, 3.3 Brand Positioning & Communications, 3.4 DMFlow — with --doc single-document runs, --view v2 re-runs, and a --combined executive stitch. Triggers on…
seo-plan
Build a 12-month SEO strategy and phased roadmap with industry templates (SaaS, ecommerce, local, publisher, agency). With fresh specialist outputs it runs as a dispatcher: scores four pillars — technical, content, topical, AI search — and makes the weakest pillar the plan's lead theme; missing specialists re-run only…
aeo-audit
Audit how a brand appears across the 6 canonical AI answer surfaces — ChatGPT, Perplexity, Google AI Mode, AI Overviews, Gemini, Copilot — probing 10-25 queries into a numbered output bundle with per-platform visibility scorecards, citation-accuracy checks, a competitor matrix, content gaps, and an optimization…
c2pa-metadata
Embed a C2PA provenance manifest into an AI-generated marketing asset (PNG, JPG, WebP, GIF, TIFF, MP4, MOV, WebM, MP3, WAV, PDF) via scripts/embed-c2pa.py — produces a signed copy of the file carrying IPTC digital-source-type AI claims, an optional c2pa.ai-disclosure assertion for EU AI Act Article 50 (applicable 2…
growth-plan
Produce the 11-section Growth Plan — the flagship Part 8 client-facing deliverable that synthesises Parts 1-7 of the engagement into a single 20-30 page executable strategy with three-scenario forecasts, an In-Market/Out-Market budget split, and a 30/60/90 timeline, exported to PDF and DOCX. Triggers on…