Files
member-console/scripts/agent-runner/scope/07-headers-rendering.md
T
cgalo5758 5829091881 Track the agent runner and its method page
The contained runner that drove the 2026-09 security audit, the README
review and four rounds of design ideation lived only inside the ignored
notebook. It moves to scripts/agent-runner/: the prepare, tools, audit,
ideation and teardown scripts, the compose and container files, the
allowlist proxy and the prompt templates, with the paths that assumed
the notebook fixed and findings written to a caller-named directory. Run
outputs, transcripts and the round-specific sheet scripts stay behind.

docs/agent-runner.md states the method: the disposable git archive
copy and the fail-closed proxy, how a task is shaped, union rather than
intersection of findings across models, separate adjudication of every
finding against the source, and the evidence a finding must carry.
2026-09-19 19:47:15 -05:00

26 lines
1.3 KiB
Markdown

# T07 — Response headers and template rendering
**Threat model.** An attacker who can get text into a page tries to execute
script in another person's browser, or to frame or downgrade the application.
**Slice.** `internal/middleware/security.go`, `internal/server/render.go`,
the error-page and static-asset paths in `internal/server/server.go`, and a
survey of `internal/embeds/templates/` for how untrusted values are emitted.
**Questions.**
1. Reproduce the CSP as served. Does every source list actually match what the
application loads today? Is any entry broader than it needs to be, and could
an attacker serve script from anything it permits?
2. Does the project's stated no-inline-script rule hold in the templates, or
are there inline handlers or inline `<script>` blocks?
3. Is any value rendered through `template.HTML`, `template.JS`, `template.URL`,
or `text/template`, and can attacker-controlled data reach it?
4. Do error pages, static assets, partial responses, and integration-mounted
routes all receive the security headers, or can a response bypass the
middleware?
5. Are the HSTS parameters appropriate for a production deployment behind TLS?
6. Does any handler write user input into a header, a redirect target, or a
cookie value without encoding?
**Not in scope.** CSRF token mechanics (T01, T03).