# Oxagen

> Steer your coding agents. Trace their work from its source to its outcome. Improve the next run. One place to steer the fleet and finish the work.

The links below provide Markdown copies of the public pages. Each copy names its source page.

## Pages

- [Oxagen | Build better software, faster.](https://oxagen.sh/index.md): Steer your coding agents. Trace their work from its source to its outcome. Improve the next run. One place to steer the fleet and finish the work.
- [Pricing](https://oxagen.sh/pricing.md): Oxagen sells operator seats and prepaid credits. Team is $99 per operator a month billed yearly or $119 billed monthly. Business is $149 billed yearly from 10 operators. Enterprise starts at $60,000 a year.
- [About Oxagen](https://oxagen.sh/about.md): Oxagen is workforce management for autonomous agents, based in Los Angeles and founded by Mac Anderson in February 2026. Contact the company here.
- [Terms of service](https://oxagen.sh/terms.md): The terms that govern your use of Oxagen: accounts, your agents, AI features, Customer Data, acceptable use, fees, and liability.
- [Privacy policy](https://oxagen.sh/privacy.md): What personal data Oxagen collects, how it uses and shares that data, how long it keeps it, and the rights you have over it.
- [Oxagen Research: the science of code graphs, agents, and self-improving systems](https://oxagen.sh/blog.md): Research notes on code graphs, AI agents, coding agents, self-improving models, self-evolving agents, and autonomous agents. Each note covers what the papers show, where the methods fail, and what it takes to govern the agents built on them.
- [Code graphs · Oxagen Research](https://oxagen.sh/blog/pillars/ontologies.md): A graph database stores each fact with its links to other facts: this function calls that one, writes this table, and changed in this commit. An agent can follow those links in a few steps. A search over text can only find passages that look like the question. Posts cover code graphs, graph retrieval, graph memory for agents, the ontology that types each fact and link, and what the research measures about each.
- [AI agents · Oxagen Research](https://oxagen.sh/blog/pillars/ai-agents.md): The science of agents: planning, tool use, memory, and evaluation. Posts run from ReAct and Reflexion to the benchmarks that measure whether an agent finishes a task, and what each run costs.
- [Coding agents · Oxagen Research](https://oxagen.sh/blog/pillars/coding-agents.md): What SWE-bench and similar benchmarks measure. How coding agents plan and repair a change across a whole repository. What a team needs before it can trust the result in production code.
- [Self-improving models · Oxagen Research](https://oxagen.sh/blog/pillars/self-improving-models.md): Research on models that learn from their own reasoning and score their own answers. Posts also cover the limits the research keeps finding, such as model collapse, where training on its own output makes a model worse, and self-correction that fails without outside feedback.
- [Self-evolving agents · Oxagen Research](https://oxagen.sh/blog/pillars/self-evolving-agents.md): Skill libraries, workflow memory, automated agent design, and code that modifies itself. Posts cover what the evidence shows and where these methods fail.
- [Autonomous agents · Oxagen Research](https://oxagen.sh/blog/pillars/autonomous-agents.md): Agents can now keep working on one task for longer. Practices that work for a ten-minute run fail over a week. Posts cover what the measurements show about long runs. They also cover what an operator has to set before a long run starts: the context the agent gets, the steering it runs under, and the record it leaves.
- [Product Releases · Oxagen Research](https://oxagen.sh/blog/pillars/product-releases.md): Release notes for each Oxagen version, with the reasons behind what shipped. Each release post links to articles that explain a feature: the problem a team hit, what the feature changes, and what it costs or limits.
- [A prepaid wallet for agent spend · Oxagen Research](https://oxagen.sh/blog/a-prepaid-wallet-for-agent-spend.md): Oxagen 5.5 prices each model call by size before it runs, pays from a monthly allowance and then a prepaid wallet, and holds new runs when the wallet is dry while started runs keep their rules.
- [Cloud sandboxes for coding agents · Oxagen Research](https://oxagen.sh/blog/cloud-sandboxes-for-coding-agents.md): A coding agent on a laptop stops when the lid closes or the plan's usage limit hits. Oxagen runs it in a sandbox on your Modal account, with every linked repository and a terminal that pauses it.
- [Decisions in one place: the Console and the Work board · Oxagen Research](https://oxagen.sh/blog/decisions-in-one-place-the-console-and-the-work-board.md): Agents raise questions, CI fails at night, and results pile up for review. The Console asks each decision as a choice, and the Work board turns a drag into the action it means.
- [Evidence an auditor can check · Oxagen Research](https://oxagen.sh/blog/evidence-an-auditor-can-check.md): An auditor should not have to trust the vendor that kept an agent's record. Oxagen 5.5.0 checks evidence offline, maps AIUC-1 control by control, and scores operators from the record.
- [Model credentials that follow the person · Oxagen Research](https://oxagen.sh/blog/model-credentials-that-follow-the-person.md): A key saved once serves every organization you work in. A run moves to your next saved login when one runs out, and a refused login asks for a new sign-in on your own machine.
- [Oxagen 5.5.0, the last beta · Oxagen Research](https://oxagen.sh/blog/oxagen-5-5-0.md): Oxagen 5.5.0 is the last beta, a pre-production release. It runs coding agents in cloud sandboxes, prices model calls before they run, and puts every decision in one place.
- [Reading a long agent run · Oxagen Research](https://oxagen.sh/blog/reading-a-long-agent-run.md): A long run hides its failures in hundreds of steps. The Run page now lists the run's loops, refusals, setup failures, and failing tools, and tells a wrapped agent when it loops.
- [Steering drafted from corrections · Oxagen Research](https://oxagen.sh/blog/steering-drafted-from-corrections.md): Oxagen now drafts steering from what people correct mid-run, reads refusals and failing tools across runs each night, and serves steering only to the agents it wraps.
- [Agents are waiting on a process built for people · Oxagen Research](https://oxagen.sh/blog/agents-are-waiting-on-a-process-built-for-people.md): An agent can write a change in minutes. Then the change waits for a person to read it. This post measures that wait, what eight companies changed about it, and how Oxagen ships at every hour.
- [Steering a run you are not watching · Oxagen Research](https://oxagen.sh/blog/steering-a-run-you-are-not-watching.md): When a long run goes wrong, most teams can stop it or type at it. Both work badly. A steer is a third option. It is a message with a delivery mode, a status, and a record.
- [The agent time horizon is doubling · Oxagen Research](https://oxagen.sh/blog/the-agent-time-horizon-is-doubling.md): METR measures how long a task an agent can finish on its own. That length has doubled about every seven months since 2019. This post covers what week-long runs mean for supervision.
- [The problem is not slop, it is your process · Oxagen Research](https://oxagen.sh/blog/the-problem-is-not-slop-it-is-your-process.md): The worry about AI slop is about output quality. The measurements point to a different cause. Teams give an agent a workflow built for people and expect it to work.
- [What an agent should be told before it starts · Oxagen Research](https://oxagen.sh/blog/what-an-agent-should-be-told-before-it-starts.md): A prompt file has no owner, no date, no scope, and no record that the agent read it. So it is a poor place for a standing rule. Steering records give each rule those things.
- [A model cannot grade its own homework · Oxagen Research](https://oxagen.sh/blog/the-limits-of-self-correction-and-model-collapse.md): Without outside feedback, self-correction fails, and a model trained on its own output gets worse. This post covers what the collapse and verifier research says to do instead.
- [Deterministic coding agents: every turn on the record · Oxagen Research](https://oxagen.sh/blog/deterministic-coding-agents-every-turn-on-the-record.md): A coding agent changed your code and no one can replay how. What the research says about feedback from running code, random sampling, and turns you can audit.
- [From STaR to DeepSeek-R1: what self-improvement means · Oxagen Research](https://oxagen.sh/blog/self-improving-models-from-star-to-self-rewarding.md): A vendor says the model improves itself. This is the research behind that claim, the signal that drives each training loop, and what stops each one.
- [Governing an agent that rewrites itself · Oxagen Research](https://oxagen.sh/blog/governing-an-agent-that-rewrites-itself.md): An agent that edits its own code needs the same review as any other change. What safety research says about reward hacking, sandboxes, oversight, and typed contracts.
- [Graph-grounded retrieval vs vector search · Oxagen Research](https://oxagen.sh/blog/graph-grounded-retrieval-vs-vector-search.md): Vector search finds the passage that looks like your question. Graph-grounded retrieval finds the fact that answers it. What the research says about the difference.
- [Self-evolving agents: what the evidence shows · Oxagen Research](https://oxagen.sh/blog/self-evolving-agents-what-the-evidence-shows.md): What changes when an agent improves itself, what checks the change, and the measured gain, across eight systems from Voyager to AlphaEvolve.
- [The science of AI agents: from ReAct to tool use · Oxagen Research](https://oxagen.sh/blog/the-science-of-ai-agents-from-react-to-tool-use.md): Planning, tool use, memory, and reflection each come from a paper that measured something. This post traces those papers and what agents still cannot do.
- [What an Ontology Buys an Agent · Oxagen Research](https://oxagen.sh/blog/what-an-ontology-buys-an-agent.md): An agent can answer with confidence from the wrong context. This post covers what classes, relations, constraints, and dated facts add to an agent's answers.
- [What SWE-bench Measures, and What It Misses · Oxagen Research](https://oxagen.sh/blog/what-swe-bench-measures-and-what-it-misses.md): Coding agents are ranked by their SWE-bench resolve rate. This post covers what that rate shows, where test-based grading goes wrong, and what the rate cannot tell you.
- [Why agents fail: measuring reliability and cost · Oxagen Research](https://oxagen.sh/blog/why-agents-fail-measuring-reliability-and-cost.md): AgentBench, WebArena, GAIA, SWE-bench, and tau-bench each measure a different thing. None of them reports what a run costs. This post covers what that hides.
- [Coding agent work for software teams | Oxagen](https://oxagen.sh/industries/software.md): Connect software work to agent runs, pull requests, checks, and costs across supported tools. Review what changed before accepting the work.
- [AI engineering by industry | Oxagen](https://oxagen.sh/industries.md): Find agent work examples for software, financial services, healthcare, and insurance teams, with the records each reviewer needs to inspect.
- [AI agent budgets and stop controls | Oxagen](https://oxagen.sh/product/mandates/budgets.md): Set spend ceilings for a workspace or an organization, pause live agent runs, and end an agent's authority. See when each control takes effect and which calls it covers.
- [AI agent identity and access | Oxagen](https://oxagen.sh/product/mandates/identity.md): Identify the agent and its operator, review access requests, and see which rule answered. Inspect credential custody for mediated connections.
- [AI agent tools and skills | Oxagen](https://oxagen.sh/product/mandates/tools-and-skills.md): Choose the tools and skills an agent may request. See how tool assignment, MCP connections, and permission checks fit the work it performs.
- [Agent permissions and controls | Oxagen](https://oxagen.sh/product/mandates.md): Set an agent's authority, budget, tools, and skills in its mandate. Review which controls apply to actions routed through Oxagen.
- [Coding agent management and work tracking | Oxagen](https://oxagen.sh/product.md): Explore how Oxagen connects work items, agent instructions, tools, and run records so your team can direct work and review its result.
- [Coding agent workflows for developers | Oxagen](https://oxagen.sh/roles/developers.md): Wrap the coding agent you already use, read the steering each run received, steer a live run, answer its questions, and follow its changes back to the run.
- [Agent access and records for security teams | Oxagen](https://oxagen.sh/roles/security.md): Inspect agent identity, access requests, and recorded decisions. Review credential custody and control scope for actions routed through Oxagen.
- [AI engineering for your role | Oxagen](https://oxagen.sh/roles.md): Find the agent work, cost, and control records your role needs, whether you lead engineering, build internal tools, review access, or ship code.
- [Live coding agent direction | Oxagen](https://oxagen.sh/product/runs/live.md): See which supported runs need a decision, send a correction, and follow its delivery status while keeping the responsible operator in view.
- [AI agent runs and work records | Oxagen](https://oxagen.sh/product/runs.md): Follow captured agent work through instructions, tool calls, code changes, and cost. Review the evidence behind each recorded result.

## Optional

- [Public pages in one file](https://oxagen.sh/llms-full.txt)
- [Sitemap](https://oxagen.sh/sitemap.xml)