Skip to content

Claude/audit docs blogs l2gsm#124

Merged
amavashev merged 4 commits intomainfrom
claude/audit-docs-blogs-L2gsm
Mar 21, 2026
Merged

Claude/audit docs blogs l2gsm#124
amavashev merged 4 commits intomainfrom
claude/audit-docs-blogs-L2gsm

Conversation

@amavashev
Copy link
Contributor

Blog Post 1: OpenAI API Budget Limits

File: blog/openai-api-budget-limits-per-user-per-run-per-tenant.md

  • ~2,700 words
  • Opens with a concrete $1,400 runaway scenario
  • Compares OpenAI's org/project/tier controls vs per-user/per-run/per-tenant (table)
  • Three patterns side by side with scope strings and pseudocode
  • Reserve-commit lifecycle explained with GPT-4o pricing math
  • Hierarchical composition and degradation strategies
  • 11 cross-links (3 blog, 5 how-to, 2 protocol, 1 quickstart)

Blog Post 2: Claude Code, Cursor, and Windsurf Budget Limits via MCP

File: blog/claude-code-cursor-windsurf-budget-limits-mcp.md

  • ~2,730 words
  • Opens with a $47 unsupervised Claude Code session
  • Frames the unsupervised session problem (table comparing MCP sessions vs traditional API calls)
  • Zero-code setup with config snippets for all four MCP hosts
  • Session-level and tool-level enforcement with a 20-step flow diagram
  • Three-way decision model with a concrete refactoring scenario
  • Wrapper vs authority argument (hallucinated token counts, ignored instructions, no atomicity)
  • 10 cross-links (3 blog, 3 how-to, 1 protocol, 3 quickstart)

claude and others added 4 commits March 21, 2026 10:27
Post 1: OpenAI API Budget Limits — search-first landing page for teams
using OpenAI who need per-user, per-run, and per-tenant budget controls.
Shows three enforcement patterns side by side and explains why org-level
provider caps are too coarse for agent workloads.

Post 2: Claude Code, Cursor, and Windsurf Budget Limits via MCP —
practical piece for developers using MCP hosts who want hard limits
without code changes. Covers zero-code setup, session-level and
tool-level enforcement, three-way decision model, and wrapper vs
authority argument.

https://claude.ai/code/session_01NEcAqmfZHsZV9WEZXNofWY
- Fix scope hierarchy diagram: workflow before agent per canonical ordering
- Fix @cycles parameter ordering: workflow before agent
- Fix caps field names: use snake_case (max_tokens) matching protocol spec

https://claude.ai/code/session_01NEcAqmfZHsZV9WEZXNofWY
update hero line, title
manifest.json still had "Hard limits on agent spend..." — updated to
match the new "Keep agents within approved spend, risk, and action
limits" messaging applied everywhere else.

https://claude.ai/code/session_01NEcAqmfZHsZV9WEZXNofWY
@amavashev amavashev merged commit f269bb1 into main Mar 21, 2026
2 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants