- Home
- Skills
- Testing & Debugging
- Stop Your AI Agent Saying "Nothing Found" Too Soon
More screenshots
Works with the AI Agents you already use
Stop Your AI Agent Saying "Nothing Found" Too Soon
An empty file search, empty test run or empty API reply often means the search was wrong, not that nothing exists. This skill runs a second, independent check before your agent reports "none found".
$5.99
Stop Your AI Agent Saying "Nothing Found" Too Soon
Example session with this skill installed
Example request
Check whether the sample folder contains text files. Do not trust one narrow glob. Use the supplied primary and fallback probes, then report the final state and count.
Command
python scripts/zero_guard.py --primary "python sample/primary_probe.py" --fallback "python sample/fallback_probe.py"
- Read your context and instructions
- Compiled the stop your ai
Example result
The primary probe checks only lowercase .txt files at the top level. It returns zero. The fallback walks nested folders and compares suffixes without case sensitivity.
PRIMARY: ZERO | exit 0 | count 0 | COUNT: 0
FALLBACK: FOUND | exit 0 | count 1 | COUNT: 1
SUMMARY: FOUND-BY-FALLBACK | count 1
The empty primary result was a false negative caused by both nesting and case.
Connects securely to your tools. The creator never sees your data.
About this skill
An empty file search, empty test run or empty API reply often means the search was wrong, not that nothing exists. This skill runs a second, independent check before your agent reports "none found".
The proof
This is a real output saved in the package, from the sample it ships with. The first check is a top-level, lowercase *.txt glob. The file it needs to find sits one folder down and is named RESULT.TXT.
PRIMARY: ZERO | exit 0 | count 0 | COUNT: 0
FALLBACK: FOUND | exit 0 | count 1 | COUNT: 1
SUMMARY: FOUND-BY-FALLBACK | count 1
The glob reported 0 and exited cleanly. The recursive, case-insensitive fallback found 1 file. Without the second check, the agent would have said "no files".
The script has 15 pytest tests. All 15 pass.
What you get
SKILL.md: the workflow your agent follows. Works in Claude Code and OpenAI Codex.RULES.md: 12 rules for empty file, text, test, API and log results.scripts/zero_guard.py: the guard. Standard-library Python (tested on 3.12), nothing to install.sample/: the narrow glob, the fallback, the nested file and the saved output above.tests/: 15 pytest tests.example-task.mdandexample-output.md: a worked example.
How to use
- Pick two checks that can fail in different ways, such as a case-sensitive glob and a recursive listing.
- Run
python scripts/zero_guard.py --primary "first command" --fallback "different command". - Read the last line:
FOUND-BY-PRIMARY: the first check found results.FOUND-BY-FALLBACK: the first check missed results the second one found.CONFIRMED-ZERO: both checks ran cleanly and returned zero.UNKNOWN: an error or timeout means you cannot conclude anything yet.
- Never turn UNKNOWN into "none found". Repair the check or use another method.
Commands run without a shell and have a timeout. Output is capped.
Known limitations
- Two checks can share the same blind spot. CONFIRMED-ZERO is only as good as the pair you choose.
- The script does not decide whether two methods are really independent. That is your call.
- It works out the count from
COUNT: N, common result phrases, pytest collection text or non-empty output lines. A tool with unrelated output can be miscounted. - It shows what two checks returned. It cannot prove that something does not exist.
FAQ
Does CONFIRMED-ZERO mean the thing definitely does not exist? No. It means two checks both ran cleanly and both found nothing. If they share a blind spot, both can be wrong.
What makes a good fallback? One that fails differently. Pair a glob with a recursive listing, an index query with a direct file check, or test collection with a look at the test files.
What happens if a command crashes or times out? The result is UNKNOWN, even if the other check returned zero.
Which agents does it work with? It is a SKILL.md skill for Claude Code and OpenAI Codex. The script is plain Python, so you can also run it by hand.
What do I need installed? Python 3 (tested on 3.12). The script uses only the standard library. The tests need pytest.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Testing & Debugging.
Counts are inferred from common CLI patterns; custom output may need manual verification. Confirmed-zero status depends entirely on the independence of the two methods chosen.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 6 days ago
- Software
Python