Add S001 integration tests; fix -n flag collisions breaking analyze/benchmark

[presidoclintonbased-alt]

Summary

#1455 — integration tests for S001/auth_gap
The issue's file pointer (tooling/sanctifier-core/src/rules/s001.rs) doesn't exist — S001 is
the AUTH_GAP finding code (finding_codes.rs), implemented by rules/auth_gap.rs (rule name
"auth_gap"). Existing SARIF/JSON integration tests in cli_tests.rs already assert "some
finding fired" against the shared vulnerable_contract.rs fixture, but none specifically target
S001/auth_gap the way the issue asks ("write tests that run the CLI against a mock workspace and
assert on the SARIF/JSON output").

Added two tests, matching every existing test in the file's conventions
(assert_cmd::Command::cargo_bin, tempdir-based fixtures):

• test_s001_auth_gap_appears_in_json_output — runs the CLI against a new, minimal fixture (a
  contract missing require_auth before a storage write, with no other issues so the assertion
  is unambiguous), asserts the JSON output's rule_violations contains a rule_name == "auth_gap" entry with a location, and that error_codes includes S001.
• test_s001_auth_gap_appears_in_sarif_output — same fixture, asserts the SARIF output's
  results array contains a ruleId == "auth_gap" entry with a physical location URI.

A real, severe pre-existing bug found by actually running these tests

Compiling passed, but running the new tests failed — sanctifier analyze panics on every
non-release invocation: main.rs's global -n/--network flag collides with analyze.rs's own
-n for --no-cache. Clap only catches short-flag collisions via a debug_assert!, so this
never shows up in a release build, but panics immediately under cargo test/cargo run — and
for anyone building this crate locally without --release. Confirmed this broke every existing
integration test in cli_tests.rs that invokes analyze, not just mine (ran
test_analyze_valid_contract before my fix; same panic, nothing to do with this PR).

Found and fixed a second instance of the identical bug: benchmark.rs's iterations field also
used -n, breaking benchmark and (via completions, which validates the whole command tree) 3
further tests. Fixed both by dropping the short flag (long form --no-cache/--iterations
unaffected) rather than reassigning --network's -n, since network is the global flag used
across every subcommand.

Test plan

• cargo test --test cli_tests: 61 passed, 0 failed, 2 ignored (the 2 ignored are
  pre-existing, unrelated — a documented flaky-HTTP-mock test). Includes both new S001 tests
  and every previously-broken analyze/benchmark/completions test, now passing.

Closes #1455
Closes #1459
Closes #1457
Closes #1454

[vercel]

@presidoclintonbased-alt is attempting to deploy a commit to the gbangbolaoluwagbemiga's projects Team on Vercel.

A member of the Team first needs to authorize it.

[drips-wave]

@presidoclintonbased-alt Great news! 🎉 Based on an automated assessment of this PR, the linked Wave issue(s) no longer count against your application limits.

You can now already apply to more issues while waiting for a review of this PR. Keep up the great work! 🚀

Learn more about application limits