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 jonZlotnik/how-to-agent --skill listen-to-nagging-doubtsgit clone --depth 1 https://github.com/jonZlotnik/how-to-agentWrote 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/jonzlotnik/how-to-agent/listen-to-nagging-doubts)<a href="https://agentmods.dev/skills/jonzlotnik/how-to-agent/listen-to-nagging-doubts"><img src="https://agentmods.dev/badge/skills/jonzlotnik/how-to-agent/listen-to-nagging-doubts/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/jonzlotnik/how-to-agent/listen-to-nagging-doubts"><img src="https://agentmods.dev/badge/skills/jonzlotnik/how-to-agent/listen-to-nagging-doubts.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 | $0.00062 | $0.00557 |
| Opus 5 | $0.00031 | $0.00279 |
| Sonnet 5 | $0.00012 | $0.00111 |
| Haiku 4.5 | $0.00006 | $0.00056 |
Grade A, and why
listen-to-nagging-doubts 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 12d 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 — 38 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Listen to Nagging Doubts
A vague unease about a piece of code is not noise. It's pattern recognition from below the threshold of articulation. The mistake is to push through it because you can't yet name it. Capture it, give it ten minutes, and either disprove it or escalate.
When to use
- "Something about this feels wrong" but you can't articulate what
- Tests pass and the code looks right, yet you don't trust it
- You reviewed your own change and felt a flinch at one spot
- You're about to merge and there's a small reluctance you'd like to ignore
Protocol
- Write the doubt down verbatim — even half-formed. "The order of these two calls is off somehow." "I don't believe this test covers the case I care about." The act of writing often sharpens it.
- Spend ≤10 minutes investigating. Read the suspicious code carefully, run the relevant test with one extra assertion, trace one input by hand. Time-box hard.
- Resolve to one of three states:
- Disproved: leave a one-line comment or commit note ("checked — handler does dedupe upstream") so future-you doesn't re-doubt.
- Confirmed: now it's a known bug; use [[fix-the-problem-not-the-blame]].
- Still vague after 10 min: escalate. File a ticket, ask a teammate, or hold the merge. Do not silently dismiss.
- Never silence a doubt without record. The next person who runs into the same instinct deserves to find your note.
Red flags (rationalizations to reject)
- "I'm probably just tired." — Maybe. Spend the 10 minutes anyway; it's cheap insurance.
- "I can't justify it, so it's not worth saying." — Doubts are evidence even before they're arguments.
- "Tests pass, doubt overruled." — Tests cover what was thought of. Doubts cover what wasn't.
Composes with
- [[fix-the-problem-not-the-blame]] — once the doubt is named, treat it like any bug investigation.
- [[prove-dont-assume]] — turning a doubt into a one-line check is exactly the prove step.
- [[broken-windows]] — a doubt about code rot is itself a signal worth acting on.
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.
- 12d ago First seen · 38 lines · 62 tokens per session scan A 0d027a3cf5f5
listen-to-nagging-doubts is a skill published in the GitHub repository jonZlotnik/how-to-agent (2 stars, last pushed 1mo ago), licensed MIT. It adds 62 tokens to every session and 557 once invoked, about $0.0003 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
systematic-debugging
4-phase root cause debugging: understand bugs before fixing.
build-test
Run the project's build / typecheck / lint / test commands and emit the build.passing + tests.passing signals devloop convergence reads.
work
Deliver one maintainer-approved EmDash issue, choosing the bug-fix path for a defect and the direct implementation path for an enhancement or task.
behavior-contract
Bug condition/postcondition formalization as testable Behavior Contracts. Defines invariants that must be preserved across fixes.
nw-bugfix
Bug fix workflow: root cause analysis → user review → regression test + fix via TDD.
test-first-bugs
Enforces a test-driven bug-fixing workflow. Use when a user reports a bug, failing code, an error, or asks to fix something.