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 kensaurus/cursor-kenji --skill housekeep-backloggit clone --depth 1 https://github.com/kensaurus/cursor-kenjiWrote 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/kensaurus/cursor-kenji/housekeep-backlog)<a href="https://agentmods.dev/skills/kensaurus/cursor-kenji/housekeep-backlog"><img src="https://agentmods.dev/badge/skills/kensaurus/cursor-kenji/housekeep-backlog.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 | $0.00077 | $0.01768 |
| Opus 5 | $0.00039 | $0.00884 |
| Sonnet 5 | $0.00015 | $0.00354 |
| Haiku 4.5 | $0.00008 | $0.00177 |
Grade A, and why
housekeep-backlog 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 4d 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 — 149 lines — stays where its author put it; the contents beside it link to each section on GitHub.
housekeep-backlog — Living register of parked work
Degree of freedom: MIXED. Discovery scan [LOW freedom — run exactly].
Classification, dedup, priority [HIGH freedom].
Apply-now for the register only. You do not implement backlog items
here. Every plan-* skill creates parked work; this is the collector.
The failure mode is not making the inventory — it goes stale within a
month and nobody can tell if debt is shrinking. So this is docs-adr
for parked work: regenerate, diff against the last run, keep stable IDs.
Prefer docs/BACKLOG.md. Use a root BACKLOG.md only if the host already
has one.
This skill vs neighbors
| Skill | Owns |
|---|---|
| housekeep-backlog (this) | Discover + dedup + write/diff the register |
complete-everything |
Execute one approved plan to closure |
burndown-full |
Drive one mechanical change to 100% |
docs-adr |
Decision memory, not parked-work inventory |
plan-stub-checker |
Stub/dead-UI plan (feeds this register) |
workflow-feature-flag |
Flag debt execution |
workflow-housekeep |
README / dead files / deps — not the backlog |
workflow-release-prep |
What's done in the dirty tree → merge-ready PR |
enhance-agent-guardrails |
Installs the same-PR / "read BACKLOG.md" wording |
How to reason (every candidate)
- Locate — file:line or doc path; form (TODO / parked phase / skip / open finding / flag)
- Interpret — what actual work is it? A bare
// TODO: fixis itself a finding — tagneeds-triage, do not drop it - Classify — tech-debt / unfinished-feature / deferred-plan-phase / test-gap / known-bug / decision-needed
- Status — still-relevant / verify-done / stale-or-obsolete
Skipping Interpret produces a register of context-free TODOs nobody can action.
Worked example
Locate:
docs/plans/auth-refactor.md— phases 1–2 checked, 3–4 unchecked; last commit 4 months ago. Interpret: phase 3 is refresh-token rotation; phase 4 is audit follow-ups parked mid-plan. Classify: deferred-plan-phase (two items). Status: ph3 still-relevant; ph4 —audit-auth-flowshas since run, so verify-done before carrying it. Rows:BL-014 | deferred-phase | auth-refactor ph3: refresh rotation | relevant | high | docs/plans/auth-refactor.mdBL-015 | deferred-phase | auth-refactor ph4: audit follow-ups | verify-done | med | docs/plans/auth-refactor.md
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.
- 4d ago First seen · 149 lines · 77 tokens per session scan A 916963fdd3d4
housekeep-backlog is a skill published in the GitHub repository kensaurus/cursor-kenji (9 stars, last pushed 10d ago), licensed MIT. It adds 77 tokens to every session and 1,768 once invoked, about $0.0004 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 2026-09-03.
Other skills, from other repositories
help-center-article-generator
Take support ticket clusters, FAQ patterns, product documentation, or feature specs and generate structured help center articles with step-by-step instructions, screenshot placeholders, troubleshooting sections, and related article links. Batch mode generates multiple articles from a ticket export. Pure reasoning…
linkedin-job-scraper
Scrapes LinkedIn job postings using the JobSpy library (python-jobspy). Use this skill whenever the user wants to find jobs on LinkedIn, search for open roles, pull job listings, build a job pipeline, source job targets for GTM research, or monitor hiring signals. Even if the user just says "find me some jobs" or…
send-sms-textbelt
Send SMS text messages to phone numbers. Use when the user asks to send a text, send an SMS, text someone, message a phone number, or send a notification via text message.
pm
When you want to manage projects across your businesses using a kanban + Eisenhower methodology. One kanban per business (whatever portfolio of businesses, projects, or initiatives you run). Tool-agnostic — connects via API/MCP to whatever PM tool each business uses (Notion, GitHub Projects, Plane, Linear, Obsidian…
contact-cache
Track all identified/contacted people across strategies. CSV-backed contact database with dedup by LinkedIn URL or email. Prevents duplicate outreach when running strategies on a recurring cadence.
give-me-tips
Explains any senpi tip in depth, including Tip: lines in the TUI. Use when the user asks about a tip, what a tipped feature does, or which tips they can see.