Context Mode (@contextmode) header
Context Mode (@contextmode) profile picture

Spend less on coding agents. Control what they do.

London, England
Joined May 2026
Photo by Context Mode (@contextmode)Photo by Context Mode (@contextmode)Photo by Context Mode (@contextmode)Photo by Context Mode (@contextmode)Photo by Context Mode (@contextmode)Photo by Context Mode (@contextmode)Photo by Context Mode (@contextmode)Photo by Context Mode (@contextmode)Photo by Context Mode (@contextmode)Photo by Context Mode (@contextmode)
Context Mode retweeted
context mode v2.2.1 is out a bug-fix release, like a tune-up for your agent's memory. after compaction your agent gets its saved state once, and recall leaves out notes from before that compaction github.com/mksglu/context-mo…
2
5
239
Context Mode retweeted
OpenClaw cut CI usage 95.5% by caching, and @boatdotdev makes CI runs faster. the agent inside CI is next. with @contextmode Gateway: fixing a failing CI run: $0.0502 to $0.0420 (22.8% less) @claude fix on a PR: $0.1113 to $0.0872 (21.6% less) using CI on boat? DM me and your Gateway CI is free contextmode.ai/solutions/ci
been so happy to help the openclaw guys with free CI on boat.dev for everyone using CI on boat, huge perf improvements (via caching and specialized infra) + limits changes coming soon
1
2
194
Context Mode retweeted
Context Mode is like a save point for your coding agent big tool output stays in a local index, so the window fills slower. after compaction, your agent picks up where it left off
Claude Code users, what do you do when the context fills up halfway through a feature? Compact, start fresh, or write a handoff file first?
2
1
5
364
Context Mode retweeted
You told your agent "never touch prod" last week. Today it ran terraform destroy. Context Mode remembers the rule, brings back last week's notes, and Protect blocks the destroy with your team's policy. contextmode.ai
1
1
2
200
Context Mode retweeted
Context Mode is a free, local, open source plugin for 11 coding agents. The agent writes a short script, the script runs in a local sandbox, and only the answer enters the context window. The full output stays in a SQLite index on your machine, so the agent can search it later. 2.0 is a full rewrite: 9 tools, hooks on all 11 agents, one database per project, and a bench you can run yourself.
5
1
1
5
2,717
Context Mode retweeted
Stop explaining your repo to your coding agent. It compacts mid-task and forgets the helper you wrote last week. Six months later it runs npm in a pnpm repo. @contextmode keeps what your agent learns. Protect blocks the force push. Pro is $20 a month.
3
2
1
3
9,287
Context Mode retweeted
LiteLLM, Portkey and Helicone are great llm gateways. none of them were built for coding agents. LiteLLM routes 100+ providers with keys and budgets. Portkey adds fallbacks, load balancing and guardrails. Helicone logs and traces every request. a coding agent needs a different layer. @contextmode Gateway is built for that: - context saving: a long Claude Code session went from $7.20 to $2.49 - Protect: 22 capabilities that check what a command actually does - recall and memory across sessions - resume a Claude Code or Codex session on another machine - works with 10 agents, no fee on tokens full comparison: contextmode.ai/docs/landscap…
13
1
10
653
Context Mode retweeted
The model writes one ES module. Our gateway runs it in a Cloudflare Dynamic Worker with no bindings, and only the value the script returns enters the conversation. The sandbox has one way out, fetch, and each fetch goes back through the gateway. How it works:
Claude Code shipped an experimental Code Mode. Instead of calling tools one at a time and reading each result, Claude writes a bit of JavaScript that runs the calls in parallel and hands back only the output it needs. So the fix for chatty agents turned out to be batching, just like fixing every N+1 query I've reviewed in 10 years. There's this one line in the docs worth reading twice: auto mode's classifier reviews the code, not each tool call. Trying it this weekend, anyone already tested it? code.claude.com/docs/en/code…
1
2
1
4
394
Context Mode retweeted
A coding agent resends its whole history on every turn. Our gateway is one Cloudflare Worker that edits that request before the model reads it. This video follows one request through it: Durable Objects, R2, Workers AI, a Dynamic Worker, and the paired benchmarks.
2
2
3
281
Context Mode retweeted
Your CI agent is not expensive because the model is smart. It is expensive because you pay for the same prompt on every pull request. Same repo, same instructions, same job, sent again and again at full price. The bill grows with every push and nobody reads the logs. Add two env lines to the job and route it through Context Mode Gateway. Same agent, same model, and each run costs 20% to 23% less across 5 measured CI jobs. contextmode.ai/solutions/ci
1
4
383
Context Mode retweeted
Unpopular opinion: an agent in CI is the most wasteful line in your AI bill. It runs on every pull request, sends nearly the same prompt each time, and you pay full price for that repeat. Nobody watches headless runs, so nobody questions the cost. Context Mode Gateway needs two env lines in the job. Same agent, same prompt, same model, and each run costs 20% to 23% less across 5 measured CI jobs. contextmode.ai/solutions/ci
8
2
17
1,274
Context Mode retweeted
We ran the same long Claude Code session on Opus 5.5 twice, with the same task. The first run cost $7.20. The second, routed through Context Mode Gateway with every feature on, cost $2.49. That is 65.5% lower. Most of the savings came from not sending the model context it did not need, and the gateway charges no fee on tokens. contextmode.ai/docs/quick-st…
1
1
146
Context Mode retweeted
thinking about sharing the @contextmode pitch deck publicly. the idea: skip the cat-and-mouse, let it get passed around, and talk directly with any pre-seed VC who's interested. would you share your deck in public, or keep it private?
1
2
172
your coding agent stack has more than a routing problem. context saving: @cmprsr, Induction, Headroom and RTK tackle context optimization with different mechanisms. gateway: @LiteLLM @PortkeyAI @vercel @Cloudflare cover model access, routing and spend controls. memory: @mem0ai @supermemory @zep_ai preserve knowledge across sessions. protect/security: @zenitysec @LassoSecurity focus on agent security and policy enforcement. session continuity: @specstoryai @GoEntire capture coding sessions and support continuing work. observability: @langfuse and LangSmith by @LangChain trace agent activity and costs. @contextmode Gateway brings context saving, recall, cross-agent memory, skills, Protect policies and spend visibility into the request path of existing coding agents. not identical products. overlapping capabilities across six categories. compare the mechanisms, coverage and trade-offs. we linked the source docs: contextmode.ai/docs/landscap…
2
1
5
350
Context Mode retweeted
Context Mode is a free, local, open source plugin for 10 coding agents. The agent writes a short script, the script runs in a local sandbox, and only the answer enters the context window. The full output stays in a SQLite index on your machine, so the agent can search it later. 2.0 is a full rewrite: 9 tools, hooks on all 10 agents, one database per project, and a bench you can run yourself. github.com/mksglu/context-mo…
9
7
4
75
6,441
Context Mode retweeted
Context Mode, a free plugin for coding agents: 709K downloads from npm, 63K GitHub installs in 14 days, 33K GitHub stars.
2
2
6
569
that's the exact problem we built session resume for in the Context Mode gateway. you pick up where you stopped, even on another machine, and with Recall the session never compacts or forgets in the first place.
Replying to @mksglu
losing the thread right when you sit back down is brutal, a fresh summary on resume would save so much rework
2
1
6
361
Context Mode retweeted
this is the most common complaint about coding agents right now. you come back to your session, and it immediately hits compaction. half your context is gone and the agent forgets what it was doing. almost every reply says the same thing: lower the compact threshold. that's not the fix. these are the only 4 steps you need to stop losing your session (no bs): 1) stop filling the window. most of it isn't your conversation, it's raw tool output. Context Mode keeps 99% of it out of context. 2) search, don't re-read. every output gets indexed locally, so the agent pulls back the one line it needs instead of the whole log. 3) survive the compaction. session state is restored right after it, so the agent picks up where it left off. 4) install once. npm i -g context-mode@next && context-mode setup free, local, no telemetry. works in Claude Code, Codex, Gemini, Copilot CLI, OpenCode and more. 32.9k stars got us to v2.
Most annoying thing in Claude Code / Codex: I get notified to come back to a session, then immediately hit compaction. Easiest fix: if it's already near ~750k (or whatever auto-compact threshold), do the compaction first, and then notify me. cc @thsottiaux @bcherny
14
1
43
4,547