remediating-with-aws-security-agent

Name: remediating-with-aws-security-agent
Author: aws

252installs190forks

How to install remediating-with-aws-security-agent?

aws/agent-toolkit-for-aws/remediating-with-aws-security-agent

$npx -y skills add aws/agent-toolkit-for-aws --skill remediating-with-aws-security-agent

Installs into the current project.

Summary

Pull AWS Security Agent findings (penetration tests and code reviews) and drive remediation. Use this whenever the user mentions Security Agent, security findings, pentest or penetration test results, code review findings, vulnerabilities found in their AWS account, "what did the

Files · 1

View on GitHub

SKILL.md

1	# Security Agent Remediation
2
3	AWS Security Agent is a frontier agent that runs on-demand penetration tests and code
4	reviews against a customer's applications and reports verified security risks. This skill
5	takes you from "I have findings somewhere in AWS" to "I'm actively fixing the most
6	important ones," while keeping the sensitive exploit detail out of source control.
7
8	The flow has four stages, and they matter in order:
9
10	1. Discover which scans exist and how the account is configured (live, read-only).
11	2. Export the findings to a local gitignored directory.
12	3. Triage the findings into a prioritized, human-readable plan.
13	4. Remediate by offering to fix the highest-risk issues.
14
15	## Why the ordering and the guardrails matter
16
17	Findings contain working attack scripts, reproduction steps, file paths, and sometimes
18	leaked secrets or environment details. If that lands in a Git repo, a customer can
19	accidentally commit and publish a step-by-step exploit for their own production system.
20	So the non-negotiable rule is: **findings are written only to `.security-agent/`, and that
21	path is gitignored before anything is written.**
22
23	## Stage 1: Discover scans (live, read-only)
24
25	Find out what the account has. All commands are read-only `list-*` operations.
26
27	AWS Security Agent organizes data as a hierarchy — work down it:
28
29	```
30	Application (account + Region)
31	└── Agent Space (workspace for design review, code review, and pentests)
32	├── Penetration test → Pentest job → Findings
33	└── Code review → Code review job → Findings
34	```
35
36	Run these to orient yourself and show the user what exists:
37
38	```bash
39	aws securityagent list-agent-spaces
40	aws securityagent list-pentests --agent-space-id <as-...>
41	aws securityagent list-code-reviews --agent-space-id <as-...>
42	aws securityagent list-pentest-jobs-for-pentest --agent-space-id <as-...> --pentest-id <pt-...>
43	aws securityagent list-code-review-jobs-for-code-review --agent-space-id <as-...> --code-review-id <cr-...>
44	```
45
46	Job `status` is one of `IN_PROGRESS`, `STOPPING`, `STOPPED`, `FAILED`, `COMPLETED`. Only
47	`COMPLETED` jobs have a stable, full set of findings.
48
49	### Match the codebase to a scan, then confirm
50
51	Agent spaces, pentests, and code reviews are named after the application they target.
52	Before asking the user to pick from a raw list, make an informed guess about which scan
53	corresponds to this repository — the user is working in a codebase for a reason, and
54	the relevant findings are almost always for the app in front of them.
55
56	Infer the app identity from the workspace using cheap, high-signal sources:
57
58	- The repository / root directory name and the Git remote URL (`git remote -v`).
59	- Project manifests and their `name`/`description` (`package.json`, `pyproject.toml`,
60	`*.csproj`, `go.mod`, `Cargo.toml`).
61	- README titles, product/steering docs, and any obvious product or company name.
62	- Distinctive frameworks or domains that match a scan title.
63
64	Compare those signals against the agent space / scan names (case-insensitive, allow
65	partial and fuzzy matches).
66	Then always confirm before exporting — present your best guess and your reasoning, and
67	let the user correct it:
68
69	> "This repo looks like `<product>` (from `<signal>`), which matches the <name> agent
70	> space. Use that, or pick another? [Other Agent Space names, ...]"
71
72	If nothing matches with reasonable confidence, say so plainly and show the full list rather
73	than forcing a wrong guess. Never export from a guessed scan without the user's confirmation.
74
75	## Stage 2: Export findings to `.security-agent/` (gitignored)
76
77	Pull findings using AWS CLI commands. Write everything into `.security-agent/` in the repo —
78	never to chat or stdout — because findings include working attack scripts, reproduction
79	steps, and sometimes leaked secrets.
80
81	### 1. Lock down the output directory before pulling anything
82
83	```bash
84	mkdir -p .security-agent
85	echo '*' > .security-agent/.gitignore
86	```
87
88	### 2. Resolve the latest COMPLETED job
89
90	You should already have the `agentSpaceId` and the pentest/code-review id from Stage

People who liked this also viewed

6 picks

Type

remediating-with-aws-security-agent

252installs190forks

View on GitHub

How to install remediating-with-aws-security-agent?

aws/agent-toolkit-for-aws/remediating-with-aws-security-agent

$npx -y skills add aws/agent-toolkit-for-aws --skill remediating-with-aws-security-agent

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Security Agent Remediation
2
3	AWS Security Agent is a frontier agent that runs on-demand penetration tests and code
4	reviews against a customer's applications and reports verified security risks. This skill
5	takes you from "I have findings somewhere in AWS" to "I'm actively fixing the most
6	important ones," while keeping the sensitive exploit detail out of source control.
7
8	The flow has four stages, and they matter in order:
9
10	1. Discover which scans exist and how the account is configured (live, read-only).
11	2. Export the findings to a local gitignored directory.
12	3. Triage the findings into a prioritized, human-readable plan.
13	4. Remediate by offering to fix the highest-risk issues.
14
15	## Why the ordering and the guardrails matter
16
17	Findings contain working attack scripts, reproduction steps, file paths, and sometimes
18	leaked secrets or environment details. If that lands in a Git repo, a customer can
19	accidentally commit and publish a step-by-step exploit for their own production system.
20	So the non-negotiable rule is: **findings are written only to `.security-agent/`, and that
21	path is gitignored before anything is written.**
22
23	## Stage 1: Discover scans (live, read-only)
24
25	Find out what the account has. All commands are read-only `list-*` operations.
26
27	AWS Security Agent organizes data as a hierarchy — work down it:
28
29	```
30	Application (account + Region)
31	└── Agent Space (workspace for design review, code review, and pentests)
32	├── Penetration test → Pentest job → Findings
33	└── Code review → Code review job → Findings
34	```
35
36	Run these to orient yourself and show the user what exists:
37
38	```bash
39	aws securityagent list-agent-spaces
40	aws securityagent list-pentests --agent-space-id <as-...>
41	aws securityagent list-code-reviews --agent-space-id <as-...>
42	aws securityagent list-pentest-jobs-for-pentest --agent-space-id <as-...> --pentest-id <pt-...>
43	aws securityagent list-code-review-jobs-for-code-review --agent-space-id <as-...> --code-review-id <cr-...>
44	```
45
46	Job `status` is one of `IN_PROGRESS`, `STOPPING`, `STOPPED`, `FAILED`, `COMPLETED`. Only
47	`COMPLETED` jobs have a stable, full set of findings.
48
49	### Match the codebase to a scan, then confirm
50
51	Agent spaces, pentests, and code reviews are named after the application they target.
52	Before asking the user to pick from a raw list, make an informed guess about which scan
53	corresponds to this repository — the user is working in a codebase for a reason, and
54	the relevant findings are almost always for the app in front of them.
55
56	Infer the app identity from the workspace using cheap, high-signal sources:
57
58	- The repository / root directory name and the Git remote URL (`git remote -v`).
59	- Project manifests and their `name`/`description` (`package.json`, `pyproject.toml`,
60	`*.csproj`, `go.mod`, `Cargo.toml`).
61	- README titles, product/steering docs, and any obvious product or company name.
62	- Distinctive frameworks or domains that match a scan title.
63
64	Compare those signals against the agent space / scan names (case-insensitive, allow
65	partial and fuzzy matches).
66	Then always confirm before exporting — present your best guess and your reasoning, and
67	let the user correct it:
68
69	> "This repo looks like `<product>` (from `<signal>`), which matches the <name> agent
70	> space. Use that, or pick another? [Other Agent Space names, ...]"
71
72	If nothing matches with reasonable confidence, say so plainly and show the full list rather
73	than forcing a wrong guess. Never export from a guessed scan without the user's confirmation.
74
75	## Stage 2: Export findings to `.security-agent/` (gitignored)
76
77	Pull findings using AWS CLI commands. Write everything into `.security-agent/` in the repo —
78	never to chat or stdout — because findings include working attack scripts, reproduction
79	steps, and sometimes leaked secrets.
80
81	### 1. Lock down the output directory before pulling anything
82
83	```bash
84	mkdir -p .security-agent
85	echo '*' > .security-agent/.gitignore
86	```
87
88	### 2. Resolve the latest COMPLETED job
89
90	You should already have the `agentSpaceId` and the pentest/code-review id from Stage

People who liked this also viewed

6 picks

Type