workshop docs: add gh-aw doc links (#4898)

Author: github-actions[bot]

workshop/side-quest-17-04-permission-escalation.md

@@ -73,7 +73,7 @@ With this setup, the agent can read data and generate output. It has no path to
 Open your workflow file. Check the `permissions:` block and answer these three questions:
 
 - [ ] Does every permission listed have a clear reason tied to your task?
-- [ ] Are there any `write` permissions that your task does not actually use?
+- [ ] Are there any `write` [permissions](https://github.github.com/gh-aw/reference/permissions/#permission-scopes) that your task does not actually use?
 - [ ] Could you replace any `write` permission with `read` and the workflow would still work?
 
 If you answered "yes" to the second or third question, remove or downgrade that permission now.

workshop/side-quest-20-01-memory-patterns.md

@@ -115,7 +115,7 @@ tools:
 | `repo-memory:` | Enables repository-backed memory for this workflow (`true` to enable). |
 
 > [!IMPORTANT]
-> `repo-memory` requires `contents: write` in your `permissions` block so the agent can commit the updated file. Add it alongside your existing permissions. This is a broader permission than `cache-memory` requires — keep the stored data small and review commits regularly.
+> `repo-memory` requires `contents: write` in your [`permissions`](https://github.github.com/gh-aw/reference/permissions/#permission-scopes) block so the agent can commit the updated file. Add it alongside your existing permissions. This is a broader permission than `cache-memory` requires — keep the stored data small and review commits regularly.
 
 Keep the stored data small — a list of IDs or a compact summary object — to avoid cluttering your commit history with large file changes.
 

workshop/side-quest-24-01-runner-infrastructure.md

@@ -62,7 +62,7 @@ network:
 
 After a successful run, the `firewall.md` artifact provides a ready-made list of every
 domain the agent contacted — share it with your security team as an allowlist baseline.
-See [Audit Reference](side-quest-25-01-audit-reference.md) for details on reading firewall logs.
+See [Audit Reference](side-quest-25-01-audit-reference.md) for details on reading [firewall](https://github.github.com/gh-aw/reference/sandbox/#awf-agent-workflow-firewall) logs.
 
 ## :white_check_mark: Checkpoint
 

workshop/side-quest-25-01-audit-reference.md

@@ -17,7 +17,7 @@
 - **Agent AIC** — total [AI Credits](https://github.github.com/gh-aw/reference/cost-management/#ai-credits-aic) consumed by the agent turn
 - **Threat-detection AIC (⌖ AIC)** — credits consumed by the [firewall](https://github.github.com/gh-aw/reference/sandbox/#awf-agent-workflow-firewall)'s threat-detection model, reported separately from agent inference
 - **[MCP tool calls](https://github.github.com/gh-aw/guides/mcps/)** — each tool the agent invoked, with any errors
-- **Threat detection verdict** — whether prompt injection, secret leak, or malicious patch was detected
+- **[Threat detection](https://github.github.com/gh-aw/reference/threat-detection/#how-it-works) verdict** — whether prompt injection, secret leak, or malicious patch was detected
 - **[Safe outputs](https://github.github.com/gh-aw/reference/safe-outputs/)** — every safe-output declaration the agent emitted
 
 ## Artifact files explained