Day 14 of 2160 minShell Walker

Day 14: Triage Replies, Duplicates and Your Personal Playbook

Taught by Mr Elite Updated

By the end of today you can

  • Classify a triage reply as duplicate, informative, not applicable or needs more information
  • Write a calm, factual reply to a request for more information
  • Explain disclosure rules and check a finding is still in scope before replying
  • Assemble prompts, checklists and notes into a one-page reusable playbook

Warm-up

Answer from memory first, then tap to check. Recalling yesterday is what makes it stick.

In what order does a triager want a report laid out?

Title, summary, steps, impact and fix.

Why must you diff an assistant's edit against your notes?

Assistants add plausible invented details, so you check every change against what you actually proved.

You send a report you are proud of. Two days later an email arrives with one line: "Thanks, this is a duplicate." Or worse, "Informative." Your stomach drops, and the temptation is to fire back an angry message or give up on the programme.

Today is about what happens after you press submit. Triage replies are a normal part of the work, not a verdict on you. I will show you how to read each type, answer calmly, stay inside the disclosure rules, and then gather everything from the last 13 days into one playbook you can reuse.

What triage replies actually mean

Triage is the first review of your report, done by a programme's security team or a platform's staff. They check three things: is it real, is it in scope, and have we seen it before. Their answer usually falls into one of a few buckets. Knowing the bucket tells you what to do next.

Duplicate
Someone else reported the same flaw first. The earlier report is the one that counts.
Informative
The behaviour is real but the team judges it low or no security risk, so there is no reward.
Not applicable
The report does not fit the rules: out of scope, not a vulnerability, or not reproducible as written.
Needs more information
The triager cannot reproduce or understand the report yet. This is good news: they are still interested.

Notice that only one bucket is an open door. A request for more information means the report is alive. Treat it as the most important email in your inbox and answer it quickly and exactly.

Triagers read a lot of reports every day, so their replies are often short and can sound cold. That is workload, not hostility. Your job is to look past the tone and find the one factual claim in the message, because that claim decides the bucket.

text
1. "This matches report #48213, closed last month."
2. "Missing header noted. No demonstrated impact."
3. "The host you tested is not listed in scope."
4. "We cannot see the redirect. Which URL did you start from?"
Four made-up triage lines. Read each one and decide its bucket before reading on.

My answers: line 1 is a duplicate, because it points at an earlier report. Line 2 is informative, because the behaviour is accepted but no risk is shown. Line 3 is not applicable, because the problem is scope. Line 4 needs more information, because it asks a direct question. If you matched all four, you are reading replies the way triagers write them.

Quick check

A triager writes: "We could not reproduce step 3. Which account were you logged in as?" What is this?

Responding calmly: duplicates, informative and not applicable

Your first reply should almost never be written in the first hour. Read the message, close it, and come back after a short break. Then work through the same four steps every time.

Here are three shapes that work. Keep them in your playbook later and change the details each time.

text
Duplicate:
Thanks for checking. Understood that this was reported earlier.
I will aim to report sooner next time.

Informative:
Thanks for the review. I should have shown the impact more
clearly. Attached is one extra step showing the data exposed.
Happy to clarify if useful.

Not applicable:
Thanks. I read the scope list again and see the host is not
listed. Could you confirm that, so I avoid it in future?
Three short reply templates. Each thanks the triager, states one fact and asks for nothing unreasonable.

Duplicates are a fact of the work. Popular programmes receive the same easy findings again and again. The useful lesson is speed and depth: report sooner, or find the version of the bug that others miss, such as a chained impact you actually proved.

Watch me do it: answer a request for more information

Here is a mock triage message about a reflected XSS report. Everything is made up and aimed at no real system. I will draft the reply, use an assistant carefully, and check the result. If you have no assistant account, write the draft by hand and skip step 3.

bash
mkdir -p ~/bounty/day14 && cd ~/bounty/day14
touch playbook.md triage-message.txt reply-notes.md reply.md
ls
Creates today's folder and four empty files, then lists them.

Now open triage-message.txt in any editor. In a terminal, nano triage-message.txt works; paste the message below, press Ctrl+O then Enter to save, and Ctrl+X to exit.

text
Triager: Thanks for your report. We tried the steps but did not see
the alert box. Which browser did you use, and was the request sent
while logged in? Also, does it work without the test account?
The mock message. Save it as day14/triage-message.txt so you have it to refer to.
text
Tidy this reply to a triager. Rules: keep it under 120 words,
be polite and factual, do NOT add any detail, version number or
claim that is not in my notes, and mark anything I did not test
as "not tested".

NOTES:
- Browser: Firefox (version in my notes)
- Logged in as test user A
- Logged out: not tested
A constrained prompt. The rules stop the assistant inventing helpful but false details.
bash
grep -n -i -E "chrome|safari|edge|[0-9]+\.[0-9]+" reply.md
Prints any line mentioning other browsers or version-like numbers. Silence means none were found.

A good result reads something like: "Thanks for looking at this. I used Firefox while logged in as test user A. I have not tested logged out, and I will do so and report back." It is short, honest and shows you will follow up.

Quick check

The assistant's draft says "I also confirmed this on Chrome 120." Your notes only mention Firefox. What do you do?

Disclosure rules and staying in scope

Disclosure means telling others about a vulnerability. Most programmes forbid public disclosure unless they agree, even after the bug is fixed. Posting a write-up, sharing screenshots or sending details to a friend can break the rules and cost you your account on the platform.

Scope
The list of hosts, apps and test types the programme allows you to touch.
Coordinated disclosure
Sharing details publicly only once the programme has fixed the issue and agreed to publication.

Policy wording varies. One programme may say "no public disclosure without written consent". Another may allow write-ups after a fix. If the policy is silent, assume the strict version and ask through the platform before you share anything.

Build your personal playbook

Over the course you have collected prompts, checklists and notes in separate places. A playbook puts the best of them on one page so you can start any hunt in minutes. Think of it as a pilot's checklist: short enough to use, specific enough to matter.

Mine has five parts: scope check, recon prompt, verification habit, report skeleton and reply rules. Yours should draw on your own folders from earlier days. First, find what you already have.

bash
cd ~/bounty
ls
grep -ril "prompt" . | head
Lists your day folders, then prints up to ten files that mention the word prompt. Adjust the path if your earlier days live elsewhere.

You should see folder names such as day02 or day13, then file paths like ./day13/notes.md. If grep prints nothing, your notes use different words, so just open the folders and look. Now put the template into playbook.md.

markdown
# My Bounty Playbook
## 1. Scope check
- Read policy; list in-scope hosts, banned tests, disclosure rule
## 2. Prompts (paste-ready)
- Explain-this-request prompt:
- Redaction reminder:
- Constrained edit prompt:
## 3. Verify habit
- Every AI claim gets a curl or Burp check; log rejects
## 4. Report skeleton
- Title, summary, steps, impact, fix; proved / not tested / suspected
## 5. Triage replies
- Wait, read, check notes, answer facts only
A starter template for playbook.md. Fill each line with your own best material, copied from earlier folders.

Quick check

Your playbook is 1,800 words after pasting everything in. What is the best next move?

If it did not work

Replying angrily or instantly to a closure

Wait an hour, reread the policy and your notes, then answer with facts or not at all.

Guessing answers to look helpful

Write "not tested" for anything you did not do, then test it if it is in scope.

Posting a write-up of a report before permission

Check the disclosure section of the policy and ask the programme in writing first.

A playbook that is ten pages of copied notes

Keep one page. Link to detail elsewhere and keep only what you will actually use.

Recap

  • Triage replies fall into duplicate, informative, not applicable or needs more information
  • A request for more information means the report is still alive
  • Reply from your notes, state what was not tested, and never add invented detail
  • Check the policy for scope and disclosure before every reply
  • A one-page playbook turns 13 days of material into a repeatable start

Cheat sheet

Duplicate
Same flaw reported earlier; thank them and move on faster next time
Informative
Real behaviour, judged low risk; reply only with new evidence
Not applicable
Out of scope or not a vulnerability as written
Needs more info
Answer exactly what is asked, quickly
Reply rules
Wait, read, check notes, facts only, stay polite
Disclosure
No public sharing without the programme's written agreement

Before the next day

Mission: Reply and playbook

Practise answering a triager calmly, then collect your best material into one page you will use for the rest of the course and beyond.

About 35 min

You are done when: A day14 folder with reply.md and a one-page playbook.md.

Want more? Write two more short replies: one to a duplicate closure and one to a not applicable closure, and add a reply template for each to your playbook.

Quiz

4 questions. Your score is saved with your progress.

1. A report is closed as a duplicate. What is the most useful response?

2. Why mark untested items as "not tested" in a reply?

3. A triager asks you to test a host that is not in the policy scope. What do you do?

4. What makes a playbook useful rather than a pile of notes?

Finish the day

Mark the day complete to save your progress.

Next: Day 15 is the capstone: you run the whole workflow from scope to report on two labs, using the playbook you built today as your starting checklist. Coming soon.

Tell someone you are doing this:WhatsAppXLinkedIn

Your notes

Private to this device. Write what you want to remember from today.