Spaces:

minhtudragon
/

headroom_3

Build error

App Files Files Community

headroom_3 / headroom

3.85 MB

Ctrl+K

Ctrl+K

33 contributors

History: 253 commits

chopratejas's picture

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

2b7049f 6 months ago

backends
Fix tool_result ordering: no stray user message between tool_calls and tool 6 months ago
cache
Reduce proxy latency, enrich telemetry, protect Bash output, bump to 0.5.6 6 months ago
ccr
Remove LLMLingua: Kompress is the sole text compressor 6 months ago
cli
Add Anthropic url overrides via ANTHROPIC_TARGET_URL env var and --anthropic-api-url flag 6 months ago
compression
Remove LLMLingua: Kompress is the sole text compressor 6 months ago
dashboard
Fix dashboard cost savings: use model list price, not moving average 6 months ago
evals
Remove LLMLingua: Kompress is the sole text compressor 6 months ago
image
Slim core dependencies: 2.5GB → 195MB install size 7 months ago
integrations
feat: add LangGraph compress_tool_messages node for ToolMessage compression 6 months ago
learn
Fix Windows drive letter path decoding in headroom learn (fixes #69) 6 months ago
memory
feat: add live traffic learning + cross-agent memory writers (--learn flag) 6 months ago
models
Remove LLMLingua: Kompress is the sole text compressor 6 months ago
perf
Remove LLMLingua: Kompress is the sole text compressor 6 months ago
prediction
Add centralized ML model configuration 8 months ago
pricing
Slim core dependencies: 2.5GB → 195MB install size 7 months ago
providers
fix: add claude-opus-4-6 (1M context) to ANTHROPIC_CONTEXT_LIMITS 6 months ago
proxy
Refactor: extract data models to headroom/proxy/models.py (Step 1/9) 6 months ago
relevance
fix: harden edge cases, expand tool exclusions, and fix mypy errors 7 months ago
reporting
fix: harden edge cases, expand tool exclusions, and fix mypy errors 7 months ago
rtk
Fix Windows Unicode crash and Bedrock Claude 4.6 model IDs 6 months ago
storage
Add pluggable adapter hooks for CCR, Storage, and TOIN backends 7 months ago
telemetry
Fix multi-worker beacon spam, cost dashboard, upsert telemetry; bump 0.5.17 6 months ago
tokenizers
Add persistent memory system with zero-latency inline extraction 8 months ago
transforms
Kompress ONNX INT8: text compression without torch dependency 6 months ago
__init__.py
6.23 kB
Fix multi-worker beacon spam, cost dashboard, upsert telemetry; bump 0.5.17 6 months ago
cli.py

137 Bytes
Add Click-based CLI with memory management commands 8 months ago
client.py
37 kB
OpenClaw plugin fixes, telemetry fix, cost tracker improvement, tokenBudget support 6 months ago
compress.py
6.82 kB
Remove LLMLingua: Kompress is the sole text compressor 6 months ago
config.py
32.5 kB
Add transforms_summary to TransformResult and /v1/compress response 6 months ago
exceptions.py

4.43 kB
Fix all ruff lint and format errors for CI 9 months ago
hooks.py
4.56 kB
Add Compression Hooks — extension points for SaaS and advanced customization 7 months ago
parser.py
10.5 kB
Reduce proxy latency, enrich telemetry, protect Bash output, bump to 0.5.6 6 months ago
py.typed

0 Bytes
Prepare for OSS release v0.2.0 9 months ago
shared_context.py

6.54 kB
Add SharedContext for multi-agent, rewrite README, fix proxy cleanup 6 months ago
tokenizer.py

2.15 kB
Initial commit: Headroom SDK - LLM context optimization toolkit 9 months ago
utils.py
7.54 kB
Pass user query context through proxy pipeline (9 call sites) 6 months ago