fix(issue-discovery): make credibility a gate, not a per-issue multiplier

[plind-junior]

Issue-discovery credibility (solved / (solved + closed)) is used twice:

• as the eligibility gate — below MIN_ISSUE_CREDIBILITY = 0.80 you score zero (issue_discovery/scoring.py:81)
• as a multiplier on every scored issue (scan.py:586 → scan.py:593)

So a discoverer who already cleared the 0.80 gate is taxed again by the same ratio on every issue — and because the ratio is flat, a few NOT_PLANNED/duplicate closures drop an active discoverer below 0.80 and zero them out. Same failure mode as #1340, but on the issue-discovery path #1340 doesn't cover.

Fix

Make credibility a gate only — exactly the OSS direction in #1340 / #1359:

# scan.py
issue.discovery_credibility_multiplier = 1.0  # gate already enforced above

Credibility is still computed, reported, and gates eligibility; it just no longer double-charges past the gate. Each issue is scored on its own multipliers (label, time decay, review quality).

Kept in-place (no shared OSS/issue helper — that was declined in #491). The flat-ratio smoothing is left as a follow-up to track whatever #1359 lands.

Tests

test_credibility_is_gate_only_not_a_per_issue_tax: an eligible discoverer at credibility 0.80 (4 valid-solved + 1 closed) keeps a neutral 1.0 multiplier and earns untaxed score.

Full suite: 878 passed. pyright / ruff / vulture clean.

[xiao-xiao-mao]

You already have 3 open PRs in this repo. The threshold is 3, so closing this one until a maintainer can triage your existing PRs. Please don't open any more — they'll be closed too.