telemetry-pipeline

Guides work on the silken_net telemetry uplink-to-minting pipeline with documented gotchas.

23|1|Updated Feb 26, 2026
One-click install
npx skills add https://github.com/Alexey-Lukin/silken_net --skill telemetry-pipeline-alexey-lukin
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: telemetry-pipeline
Source: https://github.com/Alexey-Lukin/silken_net/tree/main/.claude/skills/telemetry-pipeline
Command: npx skills add https://github.com/Alexey-Lukin/silken_net --skill telemetry-pipeline-alexey-lukin

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Working on the silken_net Proof-of-Growth telemetry pipeline (CoAP intake, unpacker, oracle verification, SCC minting) involves non-obvious invariants — strict Sidekiq queue ordering, RANGE-partitioned TelemetryLog tables, string-backed enums, dual Float Lorenz computation — that are easy to break without documented guidance. ## Core Features & Use Cases - SSOT Routing: Points to the canonical docs (CLAUDE.md sections, 05_02 pipeline canon) instead of restating them. - Indexed Gotcha Companion: 18 numbered, append-only gotchas in gotchas.md covering DID=0 handling, oracle_status enum prefix, KENOSIS validation placement, partition pruning, and DCI/Lorenz checks. - Common Task Playbooks: Step sequences for adding a telemetry field or changing minting logic, including the 2-guard oracle chain and KYC perimeter caveats. - Use Case: Before modifying TelemetryUnpackerService or a telemetry column, consult the gotcha index to avoid regressions like silent mint-path misses or fraud false positives. ## Quick Start Ask the assistant to help you add a new telemetry field to the uplink packet and unpacker, and it will consult the pipeline gotchas before editing.

Frequently Asked Questions about telemetry-pipeline

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I add a new telemetry field to the uplink pipeline?

Follow the documented sequence: update the firmware pack format, extend TelemetryUnpackerService unpacking, add a DB migration, update the Phlex dashboard component, and revise the 05_02 canon document. Check gotchas.md first for wire-format constraints.

Why are alerts delayed when telemetry traffic is high?

Sidekiq queue drain is strict by design: the uplink queue (priority 1) drains completely before the alerts queue (priority 2). A telemetry flood therefore delays alerts intentionally, and this ordering should not be changed.

Why does the server Lorenz Z value differ from the device value?

Server and firmware must both compute Lorenz in Float; any leftover to_d or BigDecimal code is a bug since FW.7. Also verify the enqueue carries the reception timestamp, since retries across midnight UTC shift the cold-derive epoch day.

Can I add model validations to TelemetryLog?

No. TelemetryLog carries no validations by design (KENOSIS TITAN); the sensor-data check lives in TelemetryUnpackerService.valid_sensor_data?. Adding validations back to the model is an explicit regression.

How should I query the partitioned TelemetryLog table?

Use the One-Home helper TelemetryLog.partition_pruned with a closed created_at window; never hand-roll created_at equality, since ISO-8601 seconds miss microsecond columns and the mint path fails silently. Retention happens only by dropping monthly partitions.