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/murphytrueman/design-system-ops/backlog-generatornpx skills add murphytrueman/design-system-ops --skill backlog-generatorgit clone --depth 1 https://github.com/murphytrueman/design-system-opsWrote 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/murphytrueman/design-system-ops/backlog-generator)<a href="https://agentmods.dev/skills/murphytrueman/design-system-ops/backlog-generator"><img src="https://agentmods.dev/badge/skills/murphytrueman/design-system-ops/backlog-generator.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 | $0.00108 | $0.01984 |
| Opus 5 | $0.00054 | $0.00992 |
| Sonnet 5 | $0.00022 | $0.00397 |
| Haiku 4.5 | $0.00011 | $0.00198 |
Grade A, and why
backlog-generator 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 5d 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 — 131 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Backlog Generator
Context
The gap between "here are the problems" and "here is the plan for the next eight weeks" is where most audit value dies. Findings get filed, not acted on. Teams read the audit, nod, then move on because the step from "TA-07: Button component tokens not aliased correctly" to "what do we build next sprint?" is too large.
This skill bridges that gap by transforming audit output into structured work items that engineers and product managers can prioritise, estimate, and execute without needing to re-read the audit or negotiate what "fix the tokens" means. The output is sprints, not a wall of problems.
Boundaries
This skill transforms existing audit findings into work items. It does not run audits itself — if no audit output exists, run the relevant audit skill first. If the audit produced zero findings, there is no backlog to generate; confirm this with the user and stop. If the findings lack severity ratings or remediation guidance, flag the gap and produce work items with what is available, but note the missing context.
Steps
1. Accept and Parse Input
Accept input in any form: copy-paste audit report, file path, reference to prior audit skill output, or natural language description of findings.
Parse findings and extract:
- Finding ID (e.g., TA-01, COL-03)
- Severity: Critical, High, Medium, Low
- Category: Token structure, component API, documentation, tooling, process
- Current state: the problem as written in the audit
- Remediation: what the audit says should happen
Do not assume findings are complete. If a finding is vague ("tokens are broken"), ask clarifying questions before proceeding.
2. Classify Work Item Type
For each finding, assign one of these work item types:
- Bug fix: Something is broken and prevents correct usage (e.g., token alias creates circular reference, component prop ignored in doc site)
- Tech debt: Working-as-designed but the design itself is unsustainable (e.g., token hierarchy is flat, component prop not typed)
- Enhancement: New capability that enables better practice (e.g., new token tier, new component variant)
- Migration: Moving from one system to another (e.g., migrating component token refs from old to new tier)
- Documentation: Clarifying existing behaviour or adding examples
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.
- 5d ago First seen · 131 lines · 108 tokens per session scan A f63fa8bf718f
backlog-generator is a skill published in the GitHub repository murphytrueman/design-system-ops (174 stars, last pushed 13d ago), licensed MIT. It adds 108 tokens to every session and 1,984 once invoked, about $0.0005 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 2026-08-30.
Other skills, from other repositories
task-publish
Files one task record as a ticket in the engineering tracker, or updates the ticket already there. The ticket carries a summary and links rather than a copy of the body, and it carries the two sections that decide when the work is finished — the done conditions and the QA checklist — drafted from the spec and the…
sync
Audits whether changes finished on working and update pages actually made it into the canonical page, applies whatever did not, then archives the working copies. This is the skill that executes the apply-after-release step — audit with zero writes, then apply, then archive, each behind its own preview and go gate.…
task-sync
Reconciles a planning-side task list against the engineering tracker it mirrors. Reads both, diagnoses the ways they have drifted apart — unfiled, duplicated, wrong parent, broken link, resurrected, field mismatch — and then proposes, waits for approval, and only then writes. Nothing is written automatically. Where…
log
Writes one day's work log as a file — what moved in the tracker, what was on the calendar, what was decided, what other people said about your work, and links to the evidence. Built to run unattended on a schedule, so it never asks and never invents; a source it cannot reach is reported as missing rather than filled…
task-draft
Turns a request source — a chat thread, a page, a spoken agreement — into a task record's context table. Statements from the source are sorted into named rows, and what is fact, what is your inference, and what nobody has decided yet are labelled apart rather than blended. Creates a new record where none exists, or…
extract-design
Extract the full design language from any website URL. Produces 8 output files including AI-optimized markdown, visual HTML preview, Tailwind config, React theme, shadcn/ui theme, Figma variables, W3C design tokens, and CSS variables. Also runs WCAG accessibility scoring. Use when user says 'extract design', 'get…