agentleFS
TIKAZITIKAZI/TIKAZ-Codex-Context-Economy/context-orchestrator/SKILL.md

context-orchestrator

Skill3 starsChanged 2 months agoMITView on GitHub

Route long text, files, webpages, code, tables, images, audio, video, or long conversations between default Context Economy and explicit ON-triggered Context On-Premise while keeping one evidence and cost contract.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Default to context-economy: dependency-free, portable, conservative, and source-preserving.

    line 16
  2. Very specific:

    Use context-onpremise only when the first non-empty control word is uppercase ON followed by end-of-input, whitespace, :, or :, or when the …

    line 17
  3. Very specific:

    Use python scripts/route_context.py --text <request> when deterministic trigger verification is useful.

    line 21

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

TIKAZ Context Orchestrator

Designed, integrated, independently refactored, and continuously maintained by TIKAZ.

What it owns

This Skill is a controller, not a third compressor. It fixes the task, evidence slots, risk, and budget, chooses exactly one primary route, and requires both routes to return traceable evidence, omissions, Pending items, fallbacks, and separate quality/cost metrics.

Routing

  • Default to context-economy: dependency-free, portable, conservative, and source-preserving.
  • Use context-onpremise only when the first non-empty control word is uppercase ON followed by end-of-input, whitespace, :, or :, or when the user explicitly invokes it.
  • Strip the valid ON control prefix before passing the remaining task downstream.
  • Lowercase on, ONLY, ON-board, 请 ON 分析, and ON appearing later in a sentence do not trigger the local route.

Use python scripts/route_context.py --text <request> when deterministic trigger verification is useful. The script parses only the control prefix; it does not open files, install dependencies, start models, or access the network.

Output contract

Return route, task, budget, evidence slots, context pack, protected facts, anchors, retrieval ledger, cost ledger, omissions, visual Pending items, and fallback reason. Report Recall/Precision, protected-fact recall, estimated tokens, file bytes, image counts, latency, and downstream answer quality separately. If labels or provider telemetry do not exist, mark the metric Pending.

Validation and fallback

A smaller context with lower evidence coverage is a failure. Expand retrieval or preserve the source when required slots are missing, facts conflict, visual evidence is unchecked, or protected literals are lost.

Example

ON:把这个扫描 PDF 和视频整理成可追溯的 Context Pack,并分别报告命中、保真和估算 Token。

Limits

The controller does not install local runtimes, download models, enable network access, or broaden authorization. Route availability depends on the installed child Skill and current environment.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.