Context Claw
context-claw
v0.2.0May 25, 2026Context window management — chunking strategies, token budgets, compression, and long-context model routing
I am the librarian who refuses to let your context window bloat. Managing context is not about summarising — it is about choosing what to forget, what to compress, what to keep as a structured index, and when to trigger a retrieval step instead of carrying dead tokens. I care about retrieval-augmented compression, sliding-window vs. semantic-importance scoring, recency bias calibration in long conversations, and cross-turn entity tracking so your agent does not forget the user is "Alice" three messages later. I produce context budgets with per-chunk attribution, compaction policies with measurable recall impact, and memory-index schemas.
OpenClaw
Hermes
Claude Code
Codex
Cursor
Windsurf
Aider
OllamaPRIMARY ACTION
Unlock with ProWhen to Use
- Run local models with private workflows
- Tune inference for local hardware
- Choose effective GGUF variants
- Benchmark practical latency and quality
Compatible Frameworks
OpenClawHermesClaude CodeCodexCursorWindsurfAiderOllama8 TOOLS
Quality Gates
- no fake claims
- chunking strategy documented
- long context model supported
- compression techniques validated
4 GATES DEFINED
Expected Outputs
Native exports per tool
openclaw/AGENTS.mdopenclaw/SOUL.mdopenclaw/TOOLS.md+7 morehermes/skills/flickclaw/context-claw/SKILL.mdhermes/skills/flickclaw/context-claw/references/workflow.mdhermes/skills/flickclaw/context-claw/references/quality-gates.md+2 moreclaude-code/CLAUDE.mdclaude-code/.claude/skills/context-claw/SKILL.mdclaude-code/.claude/skills/context-claw/references/workflow.md+3 morecodex/AGENTS.mdcodex/.flickclaw/agents/context-claw/codex.mdcodex/.flickclaw/agents/context-claw/workflow.md+2 morecursor/.cursor/rules/flickclaw-context-claw.mdccursor/.cursor/rules/flickclaw-context-claw-workflow.mdccursor/.cursor/rules/flickclaw-context-claw-quality-gates.mdcwindsurf/.windsurf/rules/flickclaw-context-claw.mdwindsurf/.windsurf/rules/flickclaw-context-claw-workflow.mdwindsurf/.windsurf/rules/flickclaw-context-claw-quality-gates.mdaider/CONVENTIONS.mdaider/aider.mdaider/.aider.conf.ymlollama/Modelfileollama/system-prompt.mdollama/template.md+1 moreInstall Commands
Install the FlickClaw CLI, then select your AI agent framework below to get the correct install command.
Step 1: Install CLI (one-time)
npm install -g @flickclaw/cli@latestStep 2: Select Framework
npm exec --yes @flickclaw/cli@latest -- install context-claw --target openclawDownload as ZIP
Example Prompt
Try this prompt with Context Claw to see what it can do:
Optimize the local model setup for performance. Benchmark current config and suggest improvements for Context Window Budget Audit with Per-Chunk Source Attribution, Memory Compaction Policy with Measured Recall-Impact Matrix, Structured Memory Index Schema with Cross-Turn Entity Registry.Example Output
IllustrativeWhat a typical Context Claw report looks like:
# Context Claw — Assessment Report **Project**: ollama-deployment **Context**: a local LLM deployment running Llama 3.1 8B on consumer GPU hardware **Generated**: 2026-07-10 ## Executive Summary The Context Claw completed its review of ollama-deployment (a local LLM deployment running Llama 3.1 8B on consumer GPU hardware). 3 findings were identified with concrete remediation steps. All quality gates were verified before delivery. ## Findings | # | Severity | Area | Finding | Recommended Action | |---|----------|------|---------|-------------------| | 1 | **P1** | Performance | Inference latency spikes to 8s under concurrency | Enable continuous batching and set max_batch=4 | | 2 | **P2** | Memory | Model consumes 18GB VRAM, headroom insufficient | Switch to Q4_K_M quantization, target <12GB | | 3 | **P2** | Setup | No Modelfile with system prompt defined | Create Modelfile with role, constraints, and templates | ## Quality Gates - [✓] entity_tracking_consistency_over_20_turns - [✓] compaction_recall_impact_measured - [✓] token_budget_per_chunk_attributed ## Outputs Generated - **Context Window Budget Audit with Per-Chunk Source Attribution**: Included in the report above. - **Memory Compaction Policy with Measured Recall-Impact Matrix**: Included in the report above. - **Structured Memory Index Schema with Cross-Turn Entity Registry**: Included in the report above. - **Long-Conversation Consistency Report (20+ Turns)**: Included in the report above. ## Validation - [x] All quality gates passed (3/3) - [x] 3 findings documented with severity and remediation - [x] 4 output sections generated - [x] Evidence collected and referenced --- *This is an illustrative example output from FlickClaw. Results vary based on project context.*