test(fuzz): v3 nesting-depth / stack-safety target

[membphis]

Background

Part of the qjson fuzz harness (#63). #63's fuzz_parse_eager clamps
max_depth to 128 (to align with serde_json's default recursion limit and
avoid differential false positives at the boundary). As a result, the v1
target does not exercise:

• qjson's real default depth (1024) or configurable ceiling (4096),
• the validate_depth boundary (off-by-one: is N the last accepted or first
  rejected depth?),
• decode-side stack safety under deep nesting.

(Note: #71 fixed a stack overflow on the encode side; this is the
decode/parse side, which is unrelated and currently uncovered.)

Depends on #63.

Goal

A non-differential fuzz target focused on nesting depth: confirm qjson
rejects over-deep input without panic / UB / stack overflow, and pin the exact
boundary behavior at the default (1024) and ceiling (4096).

Scope (this issue = one PR)

• Add a fuzz target that generates deeply-nested {/[ inputs
  (e.g. length-bounded mutation of a nested skeleton).
• Assert parse rejects beyond effective_max_depth with a clean error
  (no panic, no ASAN finding).
• Cover default (1024) and ceiling (4096), and the just-below /
  just-above boundary cases.

Acceptance Criteria

• Deeply-nested inputs never cause stack overflow or an ASAN report.
• The depth boundary (accept N / reject N+1) is asserted and documented.

Affected files

• fuzz/fuzz_targets/fuzz_depth.rs (new)

Part of #63. Depends on #63.

[coderabbitai]

🔗 Related PRs

#43 - test: add cjson and simdjson fixture coverage [merged]
#51 - perf: skip redundant validation in eager decode, inline depth check [merged]
#71 - fix(encode): add depth counter to prevent stack overflow [merged]

---

📝 Issue Planner

_{Check the box below or use the @coderabbitai plan command to generate an implementation plan and prompts that you can use with your favorite coding assistant.}

• Create Plan

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!