Refactor: extract data models to headroom/proxy/models.py (Step 1/9)

Add transforms_summary to TransformResult and /v1/compress response

Merge pull request #81 from dorphalsig/main

Fix multi-worker beacon spam, cost dashboard, upsert telemetry; bump 0.5.17

Fix dashboard cost savings: use model list price, not moving average

Merge pull request #61 from gglucass/feat/persist-savings-history

Fix WebSocket proxy: forward all headers + required OpenAI-Beta header

Add Anthropic url overrides via ANTHROPIC_TARGET_URL env var and --anthropic-api-url flag

Merge remote-tracking branch 'origin/main' into feat/persist-savings-history

feat: add historical rollups and exports

Add WebSocket proxy for /v1/responses (Codex gpt-5.4+)

Add compression for OpenAI Responses API (/v1/responses)

Fix zstd streaming decompression, add zstandard dep; bump to 0.5.14

Fix streaming tool calls, compressed request bodies, beacon field names; bump to 0.5.13

Fix cost calculation: don't mix cache savings into Headroom savings

Merge branch 'main' into feat/persist-savings-history

fix: restore anthropic compression when hooks enabled

Merge pull request #60 from gglucass/feat/add-headroom-sdk

OpenClaw plugin fixes, telemetry fix, cost tracker improvement, tokenBudget support

style: fix ruff format in server.py

fix: move extract_user_query import out of conditional scope

feat: persist proxy savings history

feat: allow overriding proxy telemetry sdk via HEADROOM_SDK

Add TypeScript SDK (headroom-ai npm package)

Remove LLMLingua: Kompress is the sole text compressor

Pass user query context through proxy pipeline (9 call sites)

fix: forward upstream ratelimit headers in streaming responses

Reduce compression latency: cache serializations, eager-load all compressors, fix Magika, bump to 0.5.5

Add telemetry beacon, fix Starlette 0.41+ crash, bump to 0.5.4

fix: close TOIN feedback loop — retrievals now flow back for learning

feat: add license key validation + phone-home usage reporter

fix: memory handler overwrites system prompt when it's a list of content blocks

feat: add live traffic learning + cross-agent memory writers (--learn flag)

fix: use cache-aware pricing from LiteLLM for dashboard cost display

feat: make token_headroom the default mode + fix Gemini handler bug

fix: propagate HEADROOM_MODE through wrap CLI and drop cache write premium penalty

fix: capture Zone 1 + Zone 2 savings in token_headroom stats

feat: add clean session summary to /stats and MCP headroom_stats

feat: add compression_cache stats to /stats endpoint

feat: add token_headroom branch to OpenAI handler

feat: add token_headroom branch to Anthropic handler

feat: configure pipeline based on HEADROOM_MODE at startup

feat: add HEADROOM_MODE config to ProxyConfig

Add wrap commands for Codex/Cursor/Aider with rtk instructions, fix savings metrics

Commit message:

Fix proxy backend bugs: env vars, tool forwarding, and provider support

Introducing headroom wrap

Updating the README to reflect the right positioning for Headroom

feat: provider-aware prefix cache tracking and combined savings dashboard

fix: dashboard metrics, TTFB tracking, eager LLMLingua loading, and multi-provider consistency