fix(mcp): honesty signals for zero-exact search and filename-only get_symbol

[RaghavChamadiya]

What

Two MCP tool-surface fixes so an agent gets an honest signal instead of a confident-but-wrong answer it will read around.

search_codebase exact-match signal

An identifier-shaped query that matches no symbol exactly used to return fuzzy neighbours with nothing to distinguish them from a real hit. It now carries exact_match: true|false, plus a short note when there is no exact match, so the agent does not anchor on whatever ranked first. The score cliff already existed internally; this surfaces it.

get_symbol filename-only resolution

get_symbol("answer.py::get_answer") (a bare filename or partial path rather than the full indexed path) dead-ended on "not found". It now resolves via a suffix rung on the file path, matching any file whose path ends with that segment on the leaf name, mirroring get_context's basename ladder. A total miss returns suggestions (real path-qualified ids) the caller can retry with.

Tests

• New TestExactMatchSignal (exact/fuzzy/mode routing) in test_search.py.
• Two suffix-rescue cases in test_symbol_ambiguity_and_format.py (resolves via suffix; total miss returns suggestions).
• Full search + symbol suites pass (60 tests). ruff clean.

[repowise-bot]

✅ Health: 7.7 (unchanged)

📋 At a glance
3 files changed health · 4 hotspots touched · 4 new findings introduced · 5 co-change pairs left out · 3 files with recent fix history.

Files & modules (2)

• packages (2 files)
  • .../mcp_server/tool_search.py
  • .../mcp_server/tool_symbol.py
• tests (2 files)
  • .../mcp/test_symbol_ambiguity_and_format.py
  • .../mcp/test_search.py

⚠️ Change risk: moderate _{(riskier than 46% of this repo's commits · raw 8.8/10)}
This change's risk is driven by:

• more lines added than baseline
• more scattered than baseline

🩹 Review priority _{(files here with the most recent bug-fix history — defects cluster, so review these first)}

• .../mcp_server/tool_search.py — fixed 5× in the last ~6 months
• .../mcp/test_search.py — fixed 2× in the last ~6 months
• .../mcp_server/tool_symbol.py — fixed once in the last ~6 months

File	Score	Δ	Why
.../mcp_server/tool_search.py	2.9 → 1.7	▼ -1.1	🔻 introduced brain method, complex method, complex conditional · ✅ resolved function hotspot, io in loop
.../mcp/test_symbol_ambiguity_and_format.py	10.0 → 9.8	▼ -0.2	score drifted on re-parse — no PR-attributable biomarkers
.../mcp/test_search.py	8.4 → 8.6	▲ +0.2	score drifted on re-parse — no PR-attributable biomarkers

💡 .../mcp_server/tool_search.py: Split this function. It carries high cyclomatic complexity and many dependents — extract cohesive responsibilities into helpers so each call site sees a smaller surface area.

🔎 More signals (2)

🔥 Hotspots touched (4)

• .../mcp_server/tool_search.py — 12 commits/90d, 11 dependents · primary owner: Raghav Chamadiya (80%)
• .../mcp/test_search.py — 5 commits/90d, 1 dependents · primary owner: Raghav Chamadiya (75%)
• .../mcp/test_symbol_ambiguity_and_format.py — 1 commits/90d, 2 dependents · primary owner: Raghav Chamadiya (100%)

1 more

• .../mcp_server/tool_symbol.py — 10 commits/90d, 7 dependents · primary owner: Swati Ahuja (58%)

🔗 Hidden coupling (2 files)

• .../mcp_server/tool_search.py co-changes with these files (not in this PR):
  • .../mcp_server/tool_overview.py (9× — 🟢 routine)
  • .../mcp_server/tool_why.py (6× — 🟢 routine)
  • .../mcp_server/tool_dead_code.py (5× — 🟢 routine)
  • .../templates/claude_md.j2 (5× — 🟢 routine)
• .../mcp_server/tool_symbol.py co-changes with .../mcp_server/tool_overview.py (6× — 🟢 routine) — not in this PR.

---

📊 Full report · ⭐ Star Repowise · 📥 Install bot · Last updated 2026-07-11 02:42 UTC
_{Silence on a single PR with [skip repowise] in the title · Per-repo toggle on repowise.dev/settings?tab=bot}