gtm-architecture-template / context
CLAUDE.md2 starsChanged 2 months agoMITView on GitHub
This repository holds the context and reusable procedures for one author. The owner, audience, offer, voice and visual identity start unfilled. A page with status: stub contains prompts, not author facts. Do not invent a biography, opinion, audience, metric, offer, voice or visual identity. Use supplied material to fill it. Creator references can inform technique but do not establish the author's experiences or beliefs. Distinguish human statements, source-backed facts and proposed interpretations. Record source dates and public-use boundaries. Unclear public-use…
In this repository54 other files it gives its agents.
- AGENTS.mdAGENTS.md
- context/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- brand-reviewSkill
- brand-systemSkill
- capture-contextSkill
- content-strategySkill
- crm-presync-checkSkill
- flowchartSkill
- graphics-designerSkill
- gtm-startSkill
- launch-videoSkill
- lead-magnet-creatorSkill
- lead-sourcingSkill
- linkedin-copywriterSkill
- long-formSkill
- manim-videoSkill
- newsletter-writerSkill
- outbound-copywriterSkill
- oxygen-dedupeSkill
- oxygen-deliverabilitySkill
- oxygen-diagnosticsSkill
- oxygen-email-infraSkill
- oxygen-gtmSkill
- oxygen-knowledgeSkill
- oxygen-linkedin-marketingSkill
- oxygen-quickstartSkill
- oxygen-recipesSkill
- oxygen-sequencer-enrollSkill
- oxygen-sequencerSkill
- oxygen-table-tidySkill
- oxygen-uniboxSkill
- oxygen-workflow-authoringSkill
- personalize-ai-columnSkill
- qmdSkill
- repurpose-contentSkill
- researcherSkill
- setup-workspaceSkill
- shorts-audioSkill
- shorts-cutSkill
- shorts-editSkill
- shorts-motionSkill
- shorts-qaSkill
- signals-sweepSkill
- spin-tagsSkill
- split-test-designerSkill
- video-useSkill
- voice-calibrationSkill
- week-postsSkill
- x-copywriterSkill
- youtube-descriptionSkill
- youtube-publisherSkill
- youtube-scriptSkill
- youtube-thumbnailSkill
The file
We split this CLAUDE.md into the rules it gives an agent
2 rules. About 0 are common, 0 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 22
Start at index.md. Search with rg or the optional repo-scoped scripts/qmd.sh.
- Very specific: line 24
Use templates/page.md for short context pages with frontmatter and Source Notes.
≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Content engine: agent rules
This repository holds the context and reusable procedures for one author. The owner, audience, offer, voice and visual identity start unfilled.
Where things live
raw/: exact source material, append only. Add a dated folder and manifest for each ingest.identity/,audience/,strategy/,voice/,brand/,inspiration/: linked context synthesized from those sources.- Notion, when connected: pillars, topics, subtopics, content drafts, calendar, hooks and creator references. Read
strategy/notion-schema.md. research/: temporary research. File durable findings in the context pages.../.agents/skills/: reusable procedures shared with the whole kit.../.claude/skillspoints to the same files.output/: local working exports and review artifacts, ignored by Git.
Empty means unknown
A page with status: stub contains prompts, not author facts. Do not invent a biography, opinion, audience, metric, offer, voice or visual identity. Use supplied material to fill it. Creator references can inform technique but do not establish the author's experiences or beliefs.
Distinguish human statements, source-backed facts and proposed interpretations. Record source dates and public-use boundaries. Unclear public-use status means the material stays out of public copy until resolved.
Read, then write
Start at index.md. Search with rg or the optional repo-scoped scripts/qmd.sh; read full sources behind claims. Before drafting, read the relevant identity, audience, strategy and voice pages. Check identity/proof.md before using a metric.
Use templates/page.md for short context pages with frontmatter and Source Notes. Update index.md and append to log.md when filing durable context. Keep the context AGENTS.md and CLAUDE.md identical.
Connections and drafts
Local connection settings belong in ignored workspace.local.json and strategy/notion.local.json. Start from their example files. They store names, preferences and resource IDs, never credentials. Use the connected tool's authentication.
Fetch the configured Notion parent and relevant data sources before writing. Never infer a destination from another repository, account or search result. If Notion is unavailable, return drafts in chat; save local working files only when requested. Do not create a second posts database.
The default taxonomy scaffold has 4 pillars, 4 topics per pillar and 4 subtopics per topic. It is adjustable to the author's strategy. Subtopics live in topic page bodies. Pick one pillar and one funnel job per post. Dates, cadence, time zone, numeric formatting, hooks and calls to action come from the current brief and author context.
Visual work
Start at brand/BRAND.md. No palette, fonts, logo, signature, footer or visual kit is preapproved. If a neutral graphic is requested while the brief is empty, produce a plain unbranded draft and label that choice. Do not treat it as the author's new identity. Deliver editable source and inspect rendered output when tools permit.
Changes and delivery
Preserve existing edits and raw sources. Run scripts/wiki-lint.sh and relevant skill checks before committing. Follow the user's Git instructions.
A request for a draft does not authorize publication, scheduling or outreach. A context capture does not authorize paid collection or enrichment. Use the user's actual scope for those actions.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

