Research: python-observability
Cached research evidence for python-observability (not authority).
Purpose
Section titled “Purpose”Python observability patterns; risks of data exposure in traces/metrics. From wshobson/agents (large multi-harness marketplace, 84 plugins/156 skills; one Markdown source to 5 harnesses).
Harness Coverage
Section titled “Harness Coverage”Target agents: antigravity, claude-code, codex, crush, cursor, gemini-cli, github-copilot, grok, opencode.
Trust And Risks
Section titled “Trust And Risks”trust_tier=curated-trust-gated; status=inspect-then-install; provenance=verified-install-command; risks=Python observability patterns; risks of data exposure in traces/metrics.; policy=Install only after trust gate / inspect source/hooks/credentials/dedupe before promotion.; evidence=wshobson/agents batches (residual + patterns) in authoring catalog; upstream https://github.com/wshobson/agents (plugins/*, docs/agent-skills.md, high star count).
Install Prerequisites
Section titled “Install Prerequisites”Install: npx skills add wshobson/agents --skill python-observability -y -g -a antigravity claude-code ... status=inspect-then-install; selector=named. Review outputs for safety (esp. generators, k8s, tf, secrets).
Upstream Maintainer
Section titled “Upstream Maintainer”wshobson/agents — MIT, broad agentic patterns + quality eval (plugin-eval).
Comparable Alternatives
Section titled “Comparable Alternatives”Similar patterns in trailofbits, vercel, local skills/* or other curated (e.g. python from modern-python, k8s from hashicorp). General python observability instruction.
> Web evidence from wshobson/agents github (README, docs, example SKILLs 2026) + config notes. Evidence only.