Remove remaining prd.md/task.md references now that they're untracked

llms.txt linked both as crawler-facing docs -- dead links now that
those files aren't in the repo. CONTRIBUTING.md told contributors to
update doc/task.md, which they can't since it's gitignored. Redirected
the one still-live CHANGELOG.md pointer to doc/judge-voting-benchmark.md
instead; left the two purely historical CHANGELOG entries describing
past states as-is.

Author: parthmax2

CHANGELOG.md

@@ -105,7 +105,7 @@ All notable changes to ghostrun are documented here. The format is based on
   time, and majority voting only marginally/non-monotonically improved
   reliability there (k=1: 49.6%, k=3: 55.8%, k=5: 51.8%) — voting is a better
   *detector* of unreliable criteria (via `disagreement_rate`) than a *fix* for
-  them. Full methodology and numbers in `doc/prd.md`.
+  them. Full methodology and numbers in `doc/judge-voting-benchmark.md`.
 - **Prompt regression tracking (PRD Phase 2, Feature 4).** Runs now snapshot the
   text every assertion saw plus each verdict, so two prompt versions can be
   compared:

CONTRIBUTING.md

@@ -29,8 +29,6 @@ real Ollama judge when available and skips gracefully otherwise.
   [doc/guide/why-not-diy.md](doc/guide/why-not-diy.md) for examples) —
   "I tested it manually" is not sufficient for anything touching concurrency,
   caching, or redaction.
-- Update `doc/task.md` if you change project status or add/complete a
-  roadmap item — it's a living document, not written after the fact.
 - Update `CHANGELOG.md` under `[Unreleased]`.
 
 ## Where things live

llms.txt

@@ -38,11 +38,9 @@ vcr-langchain, and other tools in this space.
 - [Why not just ask an LLM to write this?](doc/guide/why-not-diy.md): concrete bugs found in this project's own development (thread-safety races, secret-redaction false positives, non-monotonic majority voting) as the case for a maintained package over a generated script
 - [Judge-voting benchmark](doc/judge-voting-benchmark.md): full methodology and results for whether majority-vote judge caching actually improves reliability (answer: it's a better detector of unreliable assertions than a fix for them)
 - [Comparison with other tools](doc/comparison.md): researched, sourced comparison against DeepEval, Promptfoo, Ragas, vcr-langchain, Langfuse, Braintrust, LangSmith, Giskard, and others
-- [Product spec](doc/prd.md): the full PRD, including deviations from the original plan and why
 
 ## Project status
 
-- [Task tracker](doc/task.md): what's done, what's left, and why — updated as work happens, not retrospectively
 - [Changelog](CHANGELOG.md): release notes
 
 ## Optional