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 blakecyze/kanso --skill kanso-prgit clone --depth 1 https://github.com/blakecyze/kansoWrote 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/blakecyze/kanso/kanso-pr)<a href="https://agentmods.dev/skills/blakecyze/kanso/kanso-pr"><img src="https://agentmods.dev/badge/skills/blakecyze/kanso/kanso-pr/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/blakecyze/kanso/kanso-pr"><img src="https://agentmods.dev/badge/skills/blakecyze/kanso/kanso-pr.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 | $0.00032 | $0.02022 |
| Opus 5 | $0.00016 | $0.01011 |
| Sonnet 5 | $0.00006 | $0.00404 |
| Haiku 4.5 | $0.00003 | $0.00202 |
Grade A, and why
kanso-pr scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 220 lines — stays where its author put it; the contents beside it link to each section on GitHub.
kanso-pr
Writes PR descriptions that survive Jira migrations, Slack thread rot, and six months of organisational drift. Self-contained by default. Pulls from commit history rather than re-inventing.
The principles from kanso-principles apply. The description earns its place the same way code does.
The job
A PR description has two readers: the present reviewer and the future archaeologist. The present reviewer needs enough context to approve or push back. The future archaeologist — human or AI — needs enough context to understand why without cloning the repo, reading the diff, and reverse-engineering intent.
External links rot. The description must stand on its own. Links supplement; they don't substitute.
Before writing
Gather context in this order:
- Target branch.
!git rev-parse --abbrev-ref HEADfor current,$ARGUMENTSormainas target. - Commit history on this branch.
!git log <target>...HEAD --format="%h %s%n%b%n---". These commits already contain the what and ideally the why. Pull from them. - Diff summary.
!git diff <target>...HEAD --statfor scope awareness. - Recent merged PRs for voice calibration.
!gh pr list --state merged --limit 5 --json title,body(ifghis available). Match the tone and structure of what the repo already merges. - PR template.
!find . -path '*/PULL_REQUEST_TEMPLATE*' -o -name 'pull_request_template*'. If one exists, use it as the skeleton instead of the default.
If no PR template exists and no recent PRs are findable, fall back to the universal skeleton below.
The universal skeleton
## What
<One or two sentences. What does this PR accomplish? Must be
independently meaningful — not "see title".>
## Why
<The problem being solved. What broke, what was slow, what was
impossible that should be possible. Include the impact of NOT making
this change if it's relevant.>
## How (non-obvious decisions only)
<Skip this section entirely if everything in the diff is self-evident.
Otherwise: algorithmic choices, data structure decisions, architectural
tradeoffs that aren't visible in the code. One bullet per decision.>
## Verification
<What was actually run and the result. Test command and outcome,
typecheck, lint, manual paths exercised. If something wasn't tested,
say so explicitly and why — "no integration tests exist for this
module" beats silence. Reviewers shouldn't have to guess what's
covered.>
## Risks
<Side effects, performance implications, known limitations, anything
intentionally NOT changed. Rollback plan if non-obvious.>
## References
<Issue numbers, design docs, related PRs. Short list, not a dump.>
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.
- 9d ago First seen · 220 lines · 32 tokens per session scan A 1908bcc380da
kanso-pr is a skill published in the GitHub repository blakecyze/kanso (3 stars, last pushed 6d ago), licensed MIT. It adds 32 tokens to every session and 2,022 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
review
Structured code review with parallel audit agents, confidence-scored triage, and optional auto-fix. Examines uncommitted changes, staged diffs, commit ranges, or specific paths. Produces a tiered report (MUST-FIX / RECOMMENDED / NIT) backed by evidence, then optionally applies fixes with verification.
incident
Incident response and postmortem generation from git/deploy context. When something breaks in production, this skill builds a timeline, identifies the probable cause, and generates a structured postmortem document. Flags: --since, --service, --sev, --revert, --comms, --dry-run.
worktree
Isolate work in a git worktree. Activates when the user needs branch isolation before executing a plan, wants a clean environment for a feature, is ready to finish work in an existing worktree, or wants to reclaim worktrees that accumulated and are now polluting filesystem-level measurement.
canary
Post-deploy monitoring with browser or degraded HTTP mode. Checks console errors, performance, page load. Configurable duration (1m-30m) and interval. Reports HEALTHY/DEGRADED/BROKEN. Flags: --duration, --interval, --quick, --max-errors.
release-docs
Diff-driven documentation sync after a release. Determines what source files changed, delegates changelog to zuvo:docs, updates only docs whose source changed. Flags: --dry-run, explicit range argument.
refactor
Refactor existing code by extracting helpers, splitting files, removing duplication, or untangling dependencies. Supports resumable and batch runs. Use for structural changes that preserve behavior; use zuvo:build for new features.