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 agentmods add skills/will-pagane/claude-superdev-harness/session-endnpx skills add will-pagane/claude-superdev-harness --skill session-endgit clone --depth 1 https://github.com/will-pagane/claude-superdev-harnessWrote 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/will-pagane/claude-superdev-harness/session-end)<a href="https://agentmods.dev/skills/will-pagane/claude-superdev-harness/session-end"><img src="https://agentmods.dev/badge/skills/will-pagane/claude-superdev-harness/session-end.svg" alt="Measured on agentmods" 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 | $0.00000 | $0.02148 |
| Opus 5 | $0.00000 | $0.01074 |
| Sonnet 5 | $0.00000 | $0.00430 |
| Haiku 4.5 | $0.00000 | $0.00215 |
Grade A, and why
session-end 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 yesterday.
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 — 107 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Session End
Finished branch in → merged default branch, production in sync, pendings recorded, branch and worktree gone.
Announce at start: "Using session-end to take <branch> from pushed to merged and cleaned up."
This run has two human gates (Step 0 ambiguity, and escalations). Everything between them happens without asking. That is a gate count, not a duration promise.
Authorization
The invocation authorizes, for the work already on this branch only: committing what is left, applying its migrations, deploying its edge functions, writing the pendings file, pushing, opening the PR, merging into the default branch, and deleting the branch + worktree afterwards. This is the one place where the standing "never open a PR / never merge without an explicit ask" rule is satisfied — the invocation is the ask.
Not authorized: new features, refactors, fixing unrelated findings, force-pushing, squashing, rewriting history, deleting anything outside the branch and its worktree.
Also not authorized — tree-mutating recovery. git stash with a merge or rebase in progress, git reset --hard, git rm --cached -r .. These destroy state you did not create. Observed: a git stash push -u with MERGE_HEAD live and six hand-resolved files staged, and a git rm --cached -r . && git reset --hard on a shared main checkout with no human gate. Both are escalations, not moves.
Out-of-scope arguments: refuse and hand off — do not do the cheap part anyway. Invoked with "take enough pendings to make three specs", the right answer is that /session-build owns that, and to stop. A run that refused the execution but cut the three specs regardless got the reply "pode rodar o session build em cima" — the specs were unwanted work in the wrong skill.
Human gates — only these:
- Step 0 — confirm the branch and the pendings list when either is ambiguous.
- Escalations — a red gate triaged as
regression, an unapplied migration, a semantic merge conflict, an unverified deploy, a mid-run instruction that contradicts project law. - Nothing else. Do not ask "should I continue?" between steps, and do not invent a third gate.
What ships with it
29 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.
- assets/report-template.md 4.7 KB
- dispatch-prompts.md 3.5 KB
- pendings-template.md 1.9 KB
- references/fork-contract.md 4.4 KB
- references/repo-shapes.md 3.2 KB
- references/traps.md 18 KB
- scripts/gate.ps1 3.5 KB runs code
- scripts/gate.sh 7.0 KB runs code
- steps/lane-fork-orchestrator.md 6.9 KB
- steps/step-00-inventory.md 6.6 KB
- steps/step-01-verify.md 7.8 KB
- steps/step-02-production-state.md 3.1 KB
- steps/step-04-pendings.md 7.8 KB
- steps/step-05-push-and-pr.md 5.3 KB
- steps/step-07-merge.md 5.9 KB
- steps/step-08-sync-and-cleanup.md 5.6 KB
- steps/step-10-report.md 1.5 KB
- verify/all-gates.sh 1.5 KB runs code
- verify/assert-findings.sh 5.8 KB runs code
- verify/blocks.py 2.3 KB runs code
- verify/budgets.sh 1.7 KB runs code
- verify/contains.py 281 B runs code
- verify/expect-flag.ps1 783 B runs code
- verify/expect-flag.sh 1.6 KB runs code
- verify/ledger-probe.sh 2.0 KB runs code
- verify/lib.sh 647 B runs code
- verify/prove-red.sh 2.8 KB runs code
- verify/reconstruct.sh 4.4 KB runs code
- verify/xrefs.sh 1.3 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.
- yesterday Changed · -182 lines scan B → A bb458cccdcf1
- 5d ago First seen · 289 lines · 0 tokens per session scan B dc91a852eab7
session-end is a skill published in the GitHub repository will-pagane/claude-superdev-harness (4 stars, last pushed 3d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 2,148 tokens. 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
release-engineering
Plan and verify software releases with versioning, changelogs, release branches, feature flags, canaries, migration gates, rollback, deployment checks, and release readiness. Use when preparing a release, shipping a risky PR, coordinating app/backend/database rollout, recovering from a bad deploy, or defining release…
release
Autonomously cut a Rove (@sma1lboy/rove) release end-to-end — detect the semver bump from pending changesets (flagging an upstream minor you didn't intend), run the release gates, bump/tag/push via scripts/release.sh, then poll the GitHub Actions Release workflow with gh until npm publish completes, diagnosing CI…
changelog-generator
Draft Rove release notes as Changesets. Writes user-facing entries as .changeset/.md files for @sma1lboy/rove (consumed into packages/kobe/CHANGELOG.md at release time). Use when the user asks for "changelog", "release notes", "what changed", "add a changeset", or before cutting a version. Enforces Rove's no-soft-wrap…
prepare-release
End-to-end release automation — reads commits since last tag, infers semver bump, drafts changelog, creates release PR, merges it, waits for CI green, tags, and monitors the Docker build to completion. Use when the user says: prepare release, cut a release, bump version, create release PR, ship a new version, tag a…
lov-release-via-cicd
Configure CI/CD and publish new versions for Node, Tauri, Vite, shell, and GitHub Release projects. Use when the user asks to setup release workflow, publish a new version, verify macOS signing/notarization, recover a failed GitHub Release, or mentions "release-via-cicd", "配置并发布新版", "签名", "notarize", "GitHub Release"…
dsh-plugin-publisher
Publish a validated DSH plugin package (@deepseek-ai/dsh- or @lovstudio/dsh-) to npm, git, or tarball channels and verify it loads in the DeepSeek Harness. Use when the user asks to publish, release, or ship a plugin. 触发：发布插件 / 上架插件 / release dsh 插件。.