Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add yfge/agent-harness-skills --skill runtime-evidence-and-tracinggit clone --depth 1 https://github.com/yfge/agent-harness-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/yfge/agent-harness-skills/runtime-evidence-and-tracing)<a href="https://agentmods.dev/skills/yfge/agent-harness-skills/runtime-evidence-and-tracing"><img src="https://agentmods.dev/badge/skills/yfge/agent-harness-skills/runtime-evidence-and-tracing/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/yfge/agent-harness-skills/runtime-evidence-and-tracing"><img src="https://agentmods.dev/badge/skills/yfge/agent-harness-skills/runtime-evidence-and-tracing.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00039 | $0.00758 |
| Opus 5 | $0.00019 | $0.00379 |
| Sonnet 5 | $0.00008 | $0.00152 |
| Haiku 4.5 | $0.00004 | $0.00076 |
Grade A, and why
runtime-evidence-and-tracing scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 92 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Runtime Evidence And Tracing
Overview
Make runtime validation auditable by tying observed behavior to runtime evidence through stable IDs and artifacts.
Agents should not only say they tested something; they should leave run IDs, request IDs, logs, screenshots, interaction records, or traces. For shared harness terms, see ../../references/harness-patterns.md; when evidence surfaces are absent, use references/build-when-missing.md. For neutral runtime profiles and redaction policy, see references/runtime-profile-policy.md.
When To Use
- The user asks to connect, diagnose, or design interactive validation, runtime validation, request IDs, traces, logs, metrics, or artifacts.
- A production or test-environment issue must be traced from observed behavior to runtime evidence.
- A flow with external dependencies needs classification as code failure, dependency blocker, environment issue, or data problem.
Inputs Needed
- Target flow, entry action, command, or interface.
- Existing logging, request wrapper, headers, and observability endpoints.
- Artifact directory and run ID naming preference.
Execution Order
- First: Find the request entrypoint and existing request/log/trace propagation points.
- Then: Design run IDs, request IDs, artifact bundles, and collection commands.
- Finally: Output the evidence contract and validation path, including fallback and blocker classification.
Step-by-Step Process
- Search request wrappers, service entrypoints, logs, metrics, and trace scripts.
- Confirm whether the calling surface can generate or propagate
X-Request-IDandX-Harness-Run-ID. - If no evidence contract exists, bootstrap the minimum run artifact contract from
references/build-when-missing.md. - Design
artifacts/runs/<run_id>/with manifest, summary, logs, network, screenshots, and trace files. - Define the collection order for the target flow: start, authenticate if needed, operate, wait, read artifacts, and attribute the result.
- Define blocker categories: code regression, environment unavailable, external dependency unavailable, data missing, and not evaluable.
- Explain how PRs or ledgers should reference artifacts without committing large temporary files.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 92 lines · 39 tokens per session scan A 196a32947ede
runtime-evidence-and-tracing is a skill published in the GitHub repository yfge/agent-harness-skills (10 stars, last pushed 1mo ago), licensed MIT. It adds 39 tokens to every session and 758 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
deslop
Audit or apply evidence-backed, test-first subtractive cleanup for accumulated agent-created test bloat, verification theater, and defensive or fallback bloat while preserving independent external behavior. Invoke explicitly for semantic simplification, not generic refactoring.
learning-from-experience
Turns incidents, near misses, bad handoffs, review surprises, escaped bugs, and signals from real use into lasting fixes to your safeguards. Use after something went wrong or nearly did and a future safeguard should change. Do not use during a live incident, which comes first, or to blame someone.
reporting-shared-defects
Routes a defect found in a shared or supplied artifact (a shared prompt, skill, dependency, model, eval, or template) to the downstream teams, agents, and releases that depend on it, not just a local fix. Use when a discovered defect affects others who consume the same artifact. Do not use for a defect local to your…
responding-to-incidents
Runs a live incident the stabilize-first way — name a commander, separate facts from hypotheses, prefer reversible actions, communicate on a cadence, and drive corrective actions to closure. Use when production is broken, data is at risk, or an agent action caused harm. Do not use for routine non-incident work, or as…
add-plugin
Use when installing, disabling, enabling, or troubleshooting a third-party octomux plugin — the manifest row shape, octomux plugins list/disable/enable, reading octomux doctor output, and recovering a broken boot with --safe-mode.
create-task
Use when creating an octomux task to dispatch autonomous Claude Code agents for building features, fixing bugs, or any code changes.