CEREBRO
machine-read, human-curated
Researched and written with 265,742 tokens (in 8 · out 37,240 · cache-read 50,794 · cache-create 177,700) across 4 Claude calls
Coding agents: Claude Code and context economics
- mksglu/context-mode sandboxes MCP tool output (claims 98% reduction, e.g. a 56 KB Playwright snapshot) and persists session memory across compaction, via MCP plus hooks on 17 platforms.
- v2.1.299 adds idle compaction (200K+ token conversation idle ~54 min is compacted before the 1h prompt cache expires) and
CLAUDE_CODE_DISABLE_FILE_MENTIONS(@mentions pass paths, not contents). - v2.1.300 adds
cache_creation_5m_tokensandcache_creation_1h_tokensto the OTelapi_requestevent, so you can see cache writes split by TTL; also adds modvirtualTextfor prompt decorations. - v2.1.301 adds
CLAUDE_CODE_RETRY_WATCHDOG_MAX_USAGE_LIMIT_WAIT_MSso unattended retry mode gives up on usage limits that reset too far out, plus OTel tracing guidance in theclaude-apiskill. - xerj-org/xerj is a local index over code, docs, logs and PDFs with an Elasticsearch-compatible API, claiming 40x fewer tokens than grep for agent retrieval.
Agent orchestration and infrastructure
- How Anthropic's Dynamic Workflows Coordinate Claude Agents covers "flows", Anthropic's new multiagent orchestration type for Claude Managed Agents, announced Friday and presented at AI Engineer NY.
- Redesigning source control at Anthropic describes Anthropic replacing git with an in-house VCS after agent-driven merges went from ~2,500/day (May) to 30,000+/day (September).
- RT @AnthropicAI: Creating in-house source control is the announcement of the same post: human-scale dev infrastructure became the bottleneck, and the monorepo migrated off git in five weeks.
- Quoting Geoffrey Huntley argues for enforcing agent behaviour with pre-commit hooks that fail the commit and inject a corrective prompt, rather than relying on AGENTS.md instructions that drift.
- RT @theo: skills routing by model wishes for skills that branch on model, carry mode-specific advice (e.g. avoid slow git commands in ultrafast mode), and attribute agent work in GitHub PRs.
Agent harnesses, plugins and frameworks
- obra/superpowers is a skills-based methodology that makes the agent clarify requirements before it writes code.
- affaan-m/ECC is a harness optimisation layer (skills, memory, security) spanning Claude Code, Codex, Opencode and Cursor.
- ruvnet/ruflo is a swarm and orchestration harness installable as Claude Code plugins (core, console, mods); needs Claude Code 2.1.287+.
- anthropics/claude-plugins-official is Anthropic's curated plugin directory, with a warning that it doesn't vet what plugins ship.
- b-nnett/codex-plusplus is a macOS tweak and extension system for the Codex desktop app, patching the main process outside the app bundle.
Agent clients and CLIs
- Codex 0.163.0 adds a searchable
/configthat applies settings immediately and flags which need a restart. - RT @ClaudeDevs: Claude Code desktop app improves starting sessions faster, splitting work across several, and seeing what each is doing.
- Code users in the Claude Desktop app rely on folder groups... summarises GitHub issues on desktop: folder groups, visible context and safe permission defaults.
- RT @jullerino: Orchestrator V2 ships on T3 Code's stable channel as 0.1.0 Beta, with updated mobile builds.
- Found a GitHub repo that does Claude Code's job pitches OpenCode (214K stars, open source terminal agent) as a free alternative, citing an unverified claim that Accenture capped staff at $29/month of Claude.
- simonw/llm is a CLI and Python library for prompting dozens of remote and local models.
Agent-driven media and SaaS
- heygen-com/hyperframes renders deterministic MP4 from HTML/CSS and seekable animations, with skills for coding agents.
- I built a full DaVinci Resolve-style video editor where Claude Code is the motion designer is an open-source video editor driven by Claude Code.
- Rushilo edits video by talking to Claude Code or Codex.
- Harness by Autonomous pitches a single command centre for every coding agent across every machine.
Signals
- mksglu/context-mode: Context window optimization for AI coding agents. Sandboxes tool output (98% reduction), persists session memory, and…
- v2.1.300
- v2.1.301
- v2.1.299
- obra/superpowers: An agentic skills framework & software development methodology that works.
- Harness by Autonomous
- How Anthropic’s Dynamic Workflows Coordinate Claude Agents
- anthropics/claude-plugins-official: Official, Anthropic-managed directory of high quality Claude Code Plugins.
- RT @ClaudeDevs: We've been working hard on making the Claude Code desktop app great for running lots of sessions side by side. It's now muc…
- Found a GitHub repo that does Claude Code’s job without for free. Yesterday Accenture told staff it will only pay for $29 a month of Claude…
- affaan-m/ECC: The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Cl…
- 0.163.0
- simonw/llm: Access large language models from the command-line
- Redesigning source control at Anthropic
- b-nnett/codex-plusplus: Codex++ tweak system for the Codex desktop app
- RT @AnthropicAI: New on our Engineering blog: Creating in-house source control. In May, we merged around 2,500 code changes per day. In Sep…
- ruvnet/ruflo: 🌊 The original agent harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversatio…
- Quoting Geoffrey Huntley
- Rushilo
- Code users in the Claude Desktop app rely on folder groups, visible context, safe permission defaults and more (summary based on GitHub iss…
- xerj-org/xerj: XERJ is the new way for AI to search data. Its autoindex capability activates agents to know your data without the token was…
- I built a full DaVinci Resolve-style video editor where Claude Code is the motion designer — open source
- heygen-com/hyperframes: Write HTML. Render video. Built for agents.
- RT @jullerino: Orchestrator V2 is now out on the stable channel in @t3dotcodes! We also thought it was time we drop the Alpha branding, so…
- RT @theo: @thsottiaux - Skills that route to different behaviors based on model - Skills that have recs like "don't use slow git command" w…
Sources & citations
| Source | Fetched | In briefing |
|---|---|---|
| rss | 140 | 9 |
| github | 34 | 9 |
| x | 222 | 5 |
| 25 | 2 | |
| show_hn | 40 | 0 |
| hackernews | 30 | 0 |
| github_search | 20 | 0 |
| swipe | 20 | 0 |
| yc-rfs | 13 | 0 |
| gmail | 8 | 0 |
| yc-launch | 2 | 0 |
| seed_urls | 1 | 0 |
| crackscan | 0 | 0 |
| ossinsight | 0 | 0 |
| reddit_users | 0 | 0 |
| total | 555 | 25 |