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 wan-huiyan/agent-traffic-control --skill clean-merge-lands-line-guard-on-a-value-neither-branch-predictedgit clone --depth 1 https://github.com/wan-huiyan/agent-traffic-controlWrote 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/wan-huiyan/agent-traffic-control/clean-merge-lands-line-guard-on-a-value-neither-branch-predicted)<a href="https://agentmods.dev/skills/wan-huiyan/agent-traffic-control/clean-merge-lands-line-guard-on-a-value-neither-branch-predicted"><img src="https://agentmods.dev/badge/skills/wan-huiyan/agent-traffic-control/clean-merge-lands-line-guard-on-a-value-neither-branch-predicted/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/wan-huiyan/agent-traffic-control/clean-merge-lands-line-guard-on-a-value-neither-branch-predicted"><img src="https://agentmods.dev/badge/skills/wan-huiyan/agent-traffic-control/clean-merge-lands-line-guard-on-a-value-neither-branch-predicted.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 | $0.00287 | $0.02138 |
| Opus 5.5 | $0.00115 | $0.00855 |
| Sonnet 5.5 | $0.00057 | $0.00428 |
| Haiku 4.5 | $0.00029 | $0.00214 |
Grade A, and why
clean-merge-lands-line-guard-on-a-value-neither-branch-predicted 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 16d 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 — 185 lines — stays where its author put it; the contents beside it link to each section on GitHub.
A Clean Merge Lands a Line Guard on a Value Neither Branch Predicted
Problem
A repo has guards that pin line numbers: a test asserting that handler.py:992
still reads if name not in REGISTRY:, a published page citing
handler.py:1559, a docstring citing handler.py:1582 for the line its own
example came from. They exist because the statement is stable and its address
is not.
Two branches are open. Branch A inserts 22 lines into handler.py and, being
careful, re-aims the three citations that moved — for its tree. Branch B
adds 64 lines to a different file and re-aims the same two guard files — for
its tree. Each side is right. The edits are in different lines of the same
files, so git merges them cleanly.
The merged tree is the first tree either set of numbers has ever been read against, and it can agree with neither. Nothing conflicts, no gate is red at merge time, and the guard fails later — on the base branch, and then on every branch queued behind it.
The trap is psychological rather than technical. After the merge, those guard files look freshly fixed: modified minutes ago, by a competent session, for precisely this reason. Leaving them alone reads as respecting somebody's work rather than as skipping a check.
Context / Trigger Conditions
- A merge or rebase you just completed was clean, and it touched a file that
another file cites as
path:NNN - Two parallel branches both changed the size of a commonly-cited file (inserted a block, moved a function, deleted a guard)
git log --oneline <base>..origin/main -- <cited file>shows the other side touched it too- A test name or a comment in the diff mentions a line number
- You are about to compute a new line number as
old + lines_I_added - A guard file in the merge result has a very recent mtime and a commit message saying it fixed citations
Root cause
Line-number citations are measurements about one commit, not pointers. Git merges text, so two edits to different lines of the same file combine without complaint; it has no idea that both numbers describe the same moving target in a third file, or that the target moved twice.
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.
- 16d ago First seen · 185 lines · 287 tokens per session scan A b6d1820a12f8
clean-merge-lands-line-guard-on-a-value-neither-branch-predicted is a skill published in the GitHub repository wan-huiyan/agent-traffic-control (3 stars, last pushed 11d ago), licensed MIT. It adds 287 tokens to every session and 2,138 once invoked, about $0.0011 per session on Opus 5.5. 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 2026-09-18.
Other skills, from other repositories
cloudflare-workers-testing
Comprehensive testing guide for Cloudflare Workers using Vitest and @cloudflare/vitest-pool-workers. Use for test setup, binding mocks (D1/KV/R2/DO), integration tests, or encountering test failures, mock errors, coverage issues.
precommit
Pre-commit checks — lint:fix -> build -> test.
precommit-fast
Quick pre-commit checks — lint:fix -> test.
bun-test-lifecycle
Use for test lifecycle hooks: beforeAll, afterAll, beforeEach, afterEach, fixtures, preload.
bun-test-basics
Use for bun:test syntax, assertions, describe/it, test.skip/only/each, and basic patterns.
bun-test-mocking
Use for mock functions in Bun tests, spyOn, mock.module, implementations, and test doubles.