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 jscraik/Agent-Skills --skill he-eval-reportgit clone --depth 1 https://github.com/jscraik/Agent-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/jscraik/agent-skills/he-eval-report)<a href="https://agentmods.dev/skills/jscraik/agent-skills/he-eval-report"><img src="https://agentmods.dev/badge/skills/jscraik/agent-skills/he-eval-report/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/jscraik/agent-skills/he-eval-report"><img src="https://agentmods.dev/badge/skills/jscraik/agent-skills/he-eval-report.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 | $0.00033 | $0.02519 |
| Opus 5 | $0.00016 | $0.01260 |
| Sonnet 5 | $0.00007 | $0.00504 |
| Haiku 4.5 | $0.00003 | $0.00252 |
Grade A, and why
he-eval-report 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 6d 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 — 217 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Harness Engineering Eval Report
Philosophy
Implementation is not completion. This skill writes closure proof for exactly
one approved Harness Engineering slice, with evidence for validation, drift,
side effects, traceability, generated media when relevant, and Linear closure
safety. Higher-priority instructions, command boundaries, and local AGENTS.md
guidance remain binding.
When to Use
- A completed HE slice needs closure proof before Linear issue, milestone, project, or execution-slice closure.
- The user asks for drift validation, proof linkage, source-prompt closure, or whether completion is blocked, needs rework, or safe with follow-up.
When Not to Use
- Do not use for implementation planning, code review, strategy, or reframe design; hand off to the matching HE skill.
- Do not use to close Linear, post external comments, publish, delete, approve, or update trackers. This skill may recommend after proof, not mutate external state.
- Do not recommend closure from implementation status, missing validation, source existence, or generated media prompts without persisted artifacts.
Inputs
Selected slice, source .harness/{linear,reframes,decisions,core,strategy,triage,brainstorm,spec,plan,solutions}/
artifacts, implementation diff, validation output, branch/PR evidence, Linear
identifiers, proof artifacts, generated-media cache paths or repository media
paths when media proof is part of the slice.
Outputs
Write one report at .harness/evals/YYYY-MM-DD-JSC-###-<repo>-<issue-or-milestone>-eval.md
when Linear context is known, or .harness/evals/YYYY-MM-DD-<repo>-<issue-or-milestone>-eval.md
otherwise. Include Artifact Identity frontmatter from
Plugins/harness-engineering/references/artifact-routing-contract.md and return
schema_version, evaluated slice, validation results, drift validation, proof
artifacts, closure recommendation, follow-up work, blockers, git staging
status, staged paths, source_prompt_family_status when source-prompt
closure is in scope, Codex provenance status, PR safety trace status, runtime
persistence status, coding/testing lens status, next handoff, and confidence.
Non-trivial reports also include the BLUF review surface so the closure
recommendation, blocker consequence, and next action are visible before proof
detail.
What ships with it
14 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.
- agents/openai.yaml 210 B
- references/contract.yaml 6.4 KB
- references/evals.yaml 13 KB
- references/source-prompt-preservation.md 2.1 KB
- references/task-profile.json 1.2 KB
- scripts/agentic_validity.py 941 B runs code
- scripts/report_contract.py 2.2 KB runs code
- scripts/report_fields.py 3.6 KB runs code
- scripts/report_recommendation.py 1.9 KB runs code
- scripts/report_sections.py 3.4 KB runs code
- scripts/side_effect_authorization.py 1.8 KB runs code
- scripts/side_effect_consistency.py 1.3 KB runs code
- scripts/validate_eval_report.py 5.7 KB runs code
- tests/test_validate_eval_report.py 4.6 KB runs code
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.
- 6d ago First seen · 217 lines · 33 tokens per session scan A 32b4fc2f9d71
he-eval-report is a skill published in the GitHub repository jscraik/Agent-Skills (8 stars, last pushed 8d ago), licensed Apache-2.0. It adds 33 tokens to every session and 2,519 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 2026-09-03.
Other skills, from other repositories
trace-requirement
Link story IDs from specs/release-plan.yaml + epic capsule directories to the implementing code and tests. Produces specs/TRACEABILITYLATEST.md. Use when you want to verify coverage of a release plan, audit which stories are implemented, or find "dark" stories with no code.
process-doc
Turn an operational process into a clear SOP, role model, control map, and improvement backlog.
ag-pipeline-issue
Pipeline Issue→SPEC→Build→Verify→Test. Toda GitHub Issue gera SPEC antes de implementar, e toda implementacao e verificada contra a SPEC e testada. Use when starting work on a GitHub Issue.
taskuary-setup
Walk the owner through setting Taskuary up - the AI brain, where work arrives, the operator documents, reports and workflows - by reading the install's real state and using the screens that already exist. Use when a task was opened as a Taskuary setup walkthrough.
sprint-testing
Orchestrates in-sprint manual QA per issue across Stages 1 (Planning), 2 (Execution) and 3 (Reporting). Use for user-story testing, bug retesting, and sprint-wide QA loops. Creates the PBI folder, drives session-start, runs the triage + veto + risk-score decision tree on bugs, produces the ATP + ATR + TC artifacts in…
browser-automation
Local Python-based browser automation toolkit using Playwright. Provides command-line tools for navigating, interacting with, and testing web applications without using MCP protocols. Supports clicking, typing, hovering, screenshots, content extraction, and JavaScript execution.