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 cboone/agent-harness-plugins --skill manage-repo-licensinggit clone --depth 1 https://github.com/cboone/agent-harness-pluginsWrote 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/cboone/agent-harness-plugins/manage-repo-licensing)<a href="https://agentmods.dev/skills/cboone/agent-harness-plugins/manage-repo-licensing"><img src="https://agentmods.dev/badge/skills/cboone/agent-harness-plugins/manage-repo-licensing/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/cboone/agent-harness-plugins/manage-repo-licensing"><img src="https://agentmods.dev/badge/skills/cboone/agent-harness-plugins/manage-repo-licensing.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 | $0.00166 | $0.06151 |
| Opus 5 | $0.00083 | $0.03076 |
| Sonnet 5 | $0.00033 | $0.01230 |
| Haiku 4.5 | $0.00017 | $0.00615 |
Grade A, and why
manage-repo-licensing 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 today.
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 — 367 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Manage Repo Licensing
Encode the REUSE-style mixed-license workflow that has converged across four sibling repos (strength-model, shannon-entropy, zhang-yeung-inequality, non-shannon-inequalities). The skill is a structured companion for these rollouts: it frames policy, standardizes placement, and drives verification. It is not an automated sweeper.
Core Invariants
Every run, regardless of mode, honors these rules. They are the lessons from the four prior rollouts.
- Policy is frozen before edits. Surface the license split, the fork-lineage question, and the reference-material scope, and wait for user confirmation before touching any file.
- License-text filenames are normalized first.
LICENSES/<SPDX-ID>.txtmust match the SPDX identifier byte-for-byte (case-sensitive). Fix mis-cased filenames (APACHE-2.0.txt->Apache-2.0.txt) before any annotation pass -- otherwisereuse lintwill flag every tag. - Placement is decided by file type, not by repo convention. Use the standardized matrix in
references/file-type-matrix.md. Existing divergence is drift to migrate. REUSE.tomlis the default coverage surface. Every repo ships one with at least prose, config, generated, and structural groups (plus conditional CoC and reference-material blocks). Inline SPDX is reserved for substantive source code; per-file.licensesidecars are a narrow residual. Config files (Makefile,lakefile.toml,pyproject.toml,cspell-words.txt, dotfiles, YAML, JSONC) and generated artifacts (lock files,lean-toolchain, Pandoc output, regenerated JSON/ICS) both fall inREUSE.toml, not inline or sidecar.- Rollouts advance in small, reviewable commits. Use the ladder in
references/commit-sequence.md. Do not squash. reuse lintmust be clean when done. Plus any repo-local checks (e.g.make lint,make check-proofs).- Third-party bundled reference material is covered, not excluded. It uses
LicenseRef-Reference-Material(shared canonical text) with scale-tiered placement -- per-file sidecars for small curated collections (~20 or fewer), aREUSE.tomlglob for larger ones. Per-workLicenseRef-*exceptions require explicit justification inNOTICE. - Comments live where required, nowhere else. No
NOTICE.licensesidecar. No*.md.licensesidecars or HTML-comment SPDX inside project-authored Markdown -- both are absorbed by theREUSE.tomlprose group. NOTICEis a summary, not an override. Never write wording that implies every file is under one license.- Prose that documents SPDX syntax is a hazard. Any project-authored Markdown that contains the literal string
SPDX-License-Identifierfollowed by a colon (or a similar tag) outside a REUSE-ignore region is parsed byreuseas a real declaration and usually corrupts the lint report. Wrap such passages in<!-- REUSE-IgnoreStart -->/<!-- REUSE-IgnoreEnd -->HTML comments. This applies to this skill's own documentation, to any plan document that describes the workflow, and to README sections that show SPDX examples. The code-block fence does not protect the content; REUSE scans code blocks. - Intentional deviations are data, not drift. When running against a repo that is already REUSE-compliant, compare coverage to the matrix but do not flatten deliberate policy choices. Common legitimate deviations: (a) a fork preserving upstream attribution via per-file sidecars on unmodified files, dual-copyright MIT on lockfiles, or a compound license on the README; (b) a non-default code-bucket license (for example, Apache-2.0 rather than MIT for project-authored tooling); (c) per-work
LicenseRef-*exceptions with justification inNOTICE; (d) explicit prose path lists instead of globs when auditability matters more than elasticity. Surface these as candidate cleanups and wait for user direction before normalizing. The defaults in Sections 2-6 are a starting point; once a repo has a working policy, changes should be additive and policy-preserving, not flattening.
What ships with it
7 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.
- today Changed · +1 lines a4bd0df8a971
- 10d ago First seen · 366 lines · 166 tokens per session scan A 0f33f095e0f0
manage-repo-licensing is a skill published in the GitHub repository cboone/agent-harness-plugins (2 stars, last pushed today), licensed MIT. It adds 166 tokens to every session and 6,151 once invoked, about $0.0008 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
persona-lawyer
A review guide for legal, contract, privacy, regulatory, licensing, and compliance matters. It checks work from the viewpoint of an experienced lawyer.
iap-compliance
Detects external payment processors and WebView payment flows used for digital goods and verifies StoreKit integration, enforcing Guideline 3.1.1's requirement to use Apple IAP for all digital content purchases.
sign-in-with-apple
Detects apps that offer third-party social login without offering Sign in with Apple, violating Guideline 4.8 which requires Sign in with Apple wherever third-party authentication is used.
subscription-disclosure
Detects missing or obscured subscription disclosure elements on paywall screens, enforcing Guideline 3.1.2(c) requirements for auto-renewal terms, trial end dates, and actual price.
metadata-accuracy
Detects mismatches between features claimed in App Store metadata and features actually implemented in the codebase, enforcing Guideline 2.3.1 which prohibits misleading app descriptions.
usage-description-audit
Detects missing or insufficiently descriptive NSUsageDescription keys in Info.plist that cause guaranteed App Store rejection under Guideline 5.1.1(ii).