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-workgit 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-work)<a href="https://agentmods.dev/skills/jscraik/agent-skills/he-work"><img src="https://agentmods.dev/badge/skills/jscraik/agent-skills/he-work/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-work"><img src="https://agentmods.dev/badge/skills/jscraik/agent-skills/he-work.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 | $0.00057 | $0.01554 |
| Opus 5.5 | $0.00023 | $0.00622 |
| Sonnet 5.5 | $0.00011 | $0.00311 |
| Haiku 4.5 | $0.00006 | $0.00155 |
Grade A, and why
he-work 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 1mo 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 — 123 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Harness Engineering Work
Philosophy
Ship the smallest approved slice and leave proof: changed files, exact validation, rollback, blockers, and next handoff.
When to Use
Use when execution is already approved by a plan/spec/issue, or for a tiny low-risk fix: at most two files, no auth/security/data/CI/dependency/public API/tool/skill/plugin change, no external side effects, and a known validation command.
Inputs
Approved plan/spec/issue, selected slice, branch, dirty state, PR/Linear context, validation command, and any explicit staging or tracker authority.
Outputs
Return the YAML shape below. Use blocked_reason instead of guessing when scope, authority, validation, or ownership is unclear.
Procedure
- Check state before editing:
git status --short,git branch --show-current, andtest -f <approved-plan-or-spec>. - Confirm the approved slice names the behavior, allowed files or smallest file set, validation command, and rollback. If not, stop with
blocked_reason. - Confirm the approved source includes strict boundary fields from
../../references/spec-plan-runtime-boundary-contract.md: requested depth, execution boundary, proof boundary, runtime state/resumption key, live-state freshness, external mutation boundary,coding_lens, andtesting_lens. If any required field is absent, stale, or inferred only from chat, stop before editing withblocked_authorityorblocked_source_of_truth. - If the slice touches repo memory, artifacts, goal boards, Project Brain,
Chronicle, Local Memory, vault sync, or brownfield adoption, apply
../../references/codex-native-memory-baseline.mdbefore editing. - Edit only the approved slice. Preserve unrelated dirty files and report them without staging.
- Run the focused validation command. If it is unavailable, run the nearest syntax/audit gate and mark behavior validation
blocked. - If validation fails, fix once inside scope and re-run the same command. If it still fails, stop and hand off to
he-code-reviewwith the exact failure. - Stage nothing, commit nothing, push nothing, and update no tracker unless separately authorized.
What ships with it
6 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.
- 1mo ago Changed · +24 lines · +30 tokens per session 68605cf906b5
- 1mo ago First seen · 99 lines · 27 tokens per session scan A 7bf13875c0bb
he-work is a skill published in the GitHub repository jscraik/Agent-Skills (9 stars, last pushed 4d ago), licensed Apache-2.0. It adds 57 tokens to every session and 1,554 once invoked, about $0.0002 per session on Opus 5.5. 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
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…
finding-reconciliation
Merge the independent A/B/C findings tables into one adjudicated ledger in 07-verification/documentation-verification-report.md — normalizing to the finding schema, deduplicating by location and claim, recording per-finding corroboration without downgrading single-pass findings, promoting cross-pass disagreement to…
scoring-and-release-gate
Score a documentation package against the ten-dimension weighted rubric in references/scorecard-rubric.md with a cited justification per dimension, then evaluate the nineteen conditions in references/release-gate-conditions.md and emit a binary release-ready, conditionally-ready, or not-ready verdict with…
constraint-driven-development
Use when an Android project has no written quality bar, when the user says "set up constraints" or "define our standards", when coverage, performance, or accessibility should become enforced gates instead of per-PR arguments, when an agent keeps silencing checks or skipping tests to get to green, or when a threshold…
kyzo-plan-run-story
Parent after startstory — arm paths, spawn demolition then each T#, run board Check, spawn judge, allowlist-commit on PASS / path-restore on FAIL, Final QA verdict then checkfinalqa. Not write-story or manage-board.
scaffold-dotnet-test-project
MUST USE when an existing .NET test project was excluded from a .slnf/CI solution filter, disappeared from .sln/.slnx discovery, or lost its production ProjectReference; also for requests to set up, create, reuse, add, register, include, or repair a test project. Handles "tests pass directly but CI discovers zero"…