fix(cli): harden key import and prompt policies

[UnArbosFive]

Summary

• restore raw 64-byte sr25519 private keys without truncating them into unrelated seeds
• route flag and nested-call account names through one canonical resolver, including saved multisigs
• stop inferring arbitrary string lists as account lists
• move intent-specific picker and balance prompts into a declarative policy registry
• apply the Rust formatting currently blocking the target branch

Verification

• cargo fmt --all -- --check
• cargo test -p bittensor-core keyfiles --lib (17 passed)
• cargo clippy -p bittensor-core --all-targets -- -D warnings
• ruff check sdk/python/bittensor sdk/python/tests/unit
• ruff format --check sdk/python/bittensor sdk/python/tests/unit
• pytest -q tests/unit (1079 passed, 1 skipped)

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
subtensor	Ready	Preview	Aug 5, 2026 11:25pm

[github-actions]

🛡️ AI Review — Skeptic (security review)

VERDICT: SAFE

HIGH scrutiny: the account is 34 days old with zero public repos, but has repository admin permission, substantial merged history, no author/committer mismatch, and no identified Gittensor association; fix/btcli-combined-audit -> feat/btcli-combined.

Static review found the key-import change preserves full 64-byte sr25519 private keys, while the account-resolution and prompt-policy changes remain confined to CLI behavior. No runtime, dependency, CI, or AI-review trust-boundary changes are present.

Findings

No findings.

Conclusion

No malicious behavior or security vulnerability was found in the PR delta.

---

🔍 AI Review — Auditor (domain review)

VERDICT: 👍

Gittensor association: UNKNOWN; established repository admin with substantial recent contributions. Overlap is with the parent integration PR #3052 and incidental formatting elsewhere, not competing work.

The key restoration fix preserves full 64-byte sr25519 private keys, account-name resolution is consistently centralized, and the prompt-policy refactor retains the existing behavior with focused regression coverage. This targets a feature integration branch, so no runtime spec-version bump is required.

Findings

No findings.

Conclusion

The implementation matches the substantive PR description and introduces no blocking correctness, testing, or maintainability issues.

[github-actions]

🔄 AI review updated — Skeptic: SAFE Auditor: 👍