agentleFS
ollygardenollygarden/opentelemetry-agent-skills/skills/otel-upgrade/SKILL.md

otel-upgrade

Skill126 starsChanged 4 months agoApache-2.0View on GitHub

Assess OpenTelemetry package and Collector upgrades across ecosystems. Use when choosing or validating API, SDK, instrumentation, exporter, Collector distribution, component, image, chart, Operator, OCB build, or configuration versions, including compatibility, required changes, telemetry and runtime behavior, and rollout risk.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read only the applicable reference.

    line 18
  2. Distinctive:

    Use applicable specialized skills only when they are available.

    line 28
  3. Distinctive:

    Keep the report concise and tailored to the repository.

    line 34

≈ “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

OpenTelemetry Upgrade

Find the highest stable target supported by the available repository evidence and provide concise, case-specific upgrade advice. Unless the user asks for implementation, assess the upgrade without changing project files.

Choose the workflow

Request or repository evidenceRead
Package manifests, lockfiles, APIs, SDKs, instrumentation libraries, exporters, plugins, or dependency constraintsDependency upgrade
Collector binaries, images, charts, Operators, configuration, feature gates, components, OCB manifests, builds, or runtime behaviorCollector upgrade
A custom Collector involving independently versioned or third-party modulesRead the Collector workflow first, then the dependency workflow for relevant module constraints.

Read only the applicable reference. Read both when the upgrade crosses both surfaces.

Shared instructions

  1. Inspect the repository and runtime evidence before choosing the workflow or assuming the ecosystem.
  2. Establish the exact current and candidate versions. Record wrappers, distributions, artifacts, and embedded dependencies separately rather than treating one version as proof of another.
  3. Query authoritative registries and tagged primary upstream sources for current stable releases. Record the lookup date. Treat release candidates, beta/alpha packages, nightly artifacts, and untagged source as prereleases unless that package line intentionally publishes with a prerelease identifier; do not silently replace a project's existing experimental package line with a stable package from a different line. If no stable release meets a stated requirement, assess prereleases separately and label them clearly.
  4. Distinguish the latest available stable release from the highest validated compatible candidate. Do not call a candidate validated when the relevant checks could not run.
  5. Preserve the user's working tree by validating in a temporary copy, worktree, or equivalent isolated environment. Start with the latest stable candidate. If it fails, diagnose the cause and test selected lower versions when useful; do not assume compatibility is monotonic.
  6. Connect upstream changes to this repository's actual usage and available downstream contracts. Treat unavailable consumers, environments, and exact-version coverage as unverified.
  7. Use applicable specialized skills only when they are available. Otherwise use primary upstream sources and repository evidence, and report any resulting verification gaps.

Dependency resolution, compilation, configuration acceptance, and passing tests establish different and limited kinds of evidence. State exactly what each result demonstrates.

Report

Keep the report concise and tailored to the repository.

## Recommendation

Recommend upgrading `<subject>` from `<current>` to `<target>`, remaining on `<current>`, or deferring the upgrade. Explain the evidence, effort, and risk.

## Relevant Findings

- `<Material finding>` — explain its effect on this repository and any required action.

## Proposed Upgrade Path

Give the smallest coordinated sequence of dependency, artifact, code, configuration, and rollout changes.

## Validation

State the exact candidates, artifacts, configurations, platforms, and checks covered, their results, and what those results demonstrate.

## Remaining Risk

Mention only material behavior, environments, integrations, consumers, or artifacts that could not be verified. Omit this section when no material unverified areas were identified.

Link material upstream claims to authoritative release notes, documentation, or source. If the latest available version is not the highest validated compatible version, name both and explain the blocker.

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.