Recorded case

A bilingual knowledge product with visible limits

Publish useful explanations first, then expand languages when the evidence and workflow are ready.

Reviewed 15 October 2026 · scope: verified

Recorded method and findings

English is the base for new public content, with Russian as the second language. Translations should preserve numbers, dates, caveats and source references. A low-cost translation still needs checks for completeness and meaning before publication.

Recorded method and findings

The 14 October TVIJO receipt found 417 sitemap URLs and all 62 existing English public articles included. This added sitemap entries intended to improve discovery of existing content. It does not establish search-engine indexing, ranking or additional articles created by that change.

Recorded method and findings

The public wiki provides a read-only JSON index at /wiki/index.json with dated section titles, evidence status and English/Russian links. It requires no account, sends no commands and exposes no private operator records.

Practical checklist

Maintain one factual definition across wiki, documentation and API examples. Add a source date and a section explaining when a method fails. Translate additional languages after content quality and demand justify it.

Sources & scope

TVIJO sitemap and article coverage receipt 2026-10-14
Knowledge publication gates and language policy 2026-10-15

Keep exploring

AI orchestration: from a prompt to an accepted resultA practical design for delegation, evidence and recovery.Shared memory without shared mistakesKeep facts, hypotheses and permissions distinct across agent sessions.Engineering that can show its workTVIJO connects operators, bounded agent work and knowledge that can be checked and reused.AGENTORCH: the orchestrator has a concrete jobAGENTORCH names the orchestration role: plan, assign, observe, reconcile and accept.The task packet is the interfaceA good task describes a verifiable outcome rather than an invitation to improvise without limits.AGENTJUDGE: acceptance follows evidenceA reviewer checks the task, the artifact version and each acceptance criterion.Measure cost per accepted resultSmall task benchmarks illustrate why task class and acceptance matter more than a cheap model label.Budgets and outages belong in the state machineA reliable operator view distinguishes working, waiting, failed, blocked and accounting uncertainty.Keep VECTORBASE, evaluate retrievalUse a private source-linked index to find evidence; do not treat search results as authority.SPIDER: evidence collection and useful progressThe first implementation reads an explicit project-evidence list and produces a dated operator digest without model calls.DSHREMBOT and tmux: one operator viewAuthenticated project commands, budget visibility and explicit schedule controls connect the operator to existing execution lanes.The AI worker–Hugging Face incident: learn from controlsThe incident reports support specific engineering lessons, not a controlled ranking of our system.What the supplied AI worker screenshot actually provesThe screenshot records a safeguard-triggered switch in one session; it does not explain the exact trigger. The observation is the supplied screenshot dated 15 October 2026.Training our own model: a gated research directionA proprietary model is an option to test after retrieval, prompts and routing have measurable limits.Free learning, custom projects and future challengesTurn evidence-backed engineering into clear offers and a useful developer community.Trust the sender, not a public checksumBind the complete message to an explicitly trusted authority.