Need a hand? We read every message.

Does Codag replace my LLM?

No. Codag sits in front of your LLM. Send raw logs to us, get back just the lines that matter as schema-valid JSON, then pass that to whatever agent you're already using. Claude, GPT, Gemini, local Ollama, anything.

Will my logs be used to train models?

Your raw payloads are processed transiently. We may keep a de-identified, PII-scrubbed copy of submitted logs and use it to improve and train Codag's own models (the templater and classifier) — that copy is stripped of identifiers like emails, IPs, tokens, and secrets. We do not sell or share your data, and we never send your logs to third-party LLM providers such as OpenAI or Anthropic. See the privacy policy.

How fast is the warm path?

Under 1 second per incident on cache hits, which is roughly 95% of production traffic once your services have warmed up.

Which log formats are supported?

Anything line-oriented. JSON logs, syslog, framework-specific formats (Hadoop, Spark, HDFS, K8s), unstructured app logs. The preprocessing stage handles parsing and PII redaction before templating.

How do I get started?

Install the CLI with curl -fsSL https://codag.ai/install.sh | sh. You can try it immediately with no account: codag wrap -- your-log-command. Sign in with codag auth login for the full pipeline, then run codag setup to wire the MCP server into your agent automatically. See the product page for the full setup.