coderag-ai 0.3.3__tar.gz → 0.3.4__tar.gz

{coderag_ai-0.3.3 → coderag_ai-0.3.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: coderag-ai
-Version: 0.3.3
+Version: 0.3.4
 Summary: Token-efficient, structure-aware Code Intelligence + RAG for private source repositories.
 Project-URL: Homepage, https://github.com/galacticsurfer/coderag
 Project-URL: Repository, https://github.com/galacticsurfer/coderag

{coderag_ai-0.3.3 → coderag_ai-0.3.4}/pyproject.toml

@@ -7,7 +7,7 @@ build-backend = "hatchling.build"
 # command are both `coderag`. Plain "coderag" is taken on PyPI by an unrelated project,
 # and "code-rag" is rejected as too similar (PyPI collapses -/_/. when comparing names).
 name = "coderag-ai"
-version = "0.3.3"
+version = "0.3.4"
 description = "Token-efficient, structure-aware Code Intelligence + RAG for private source repositories."
 readme = "README.md"
 requires-python = ">=3.12"

{coderag_ai-0.3.3 → coderag_ai-0.3.4}/src/coderag/api/app.py

@@ -286,6 +286,10 @@ def metrics(session: Session = Depends(get_session)) -> s.MetricsOut:
     ) or 0)
     saved_vs_files = base_sum - base_ctx
     reduction_vs_files = round(100.0 * saved_vs_files / base_sum, 1) if base_sum else 0.0
+    from coderag.core.config import get_settings
+
+    _s = get_settings()
+    pin, pout = _s.price_input_per_mtok, _s.price_output_per_mtok
     return s.MetricsOut(
         repositories=count(Repository), symbols=count(Symbol),
         embeddings=count(SymbolEmbedding), relationships=count(SymbolRelationship),
@@ -297,6 +301,10 @@ def metrics(session: Session = Depends(get_session)) -> s.MetricsOut:
         total_tokens_saved=saved, avg_token_reduction_percent=reduction,
         total_baseline_tokens=base_sum, total_saved_vs_files=saved_vs_files,
         reduction_vs_files_percent=reduction_vs_files,
+        price_input_per_mtok=pin, price_output_per_mtok=pout,
+        cost_saved_vs_files_usd=round(saved_vs_files / 1e6 * pin, 4),
+        cost_context_sent_usd=round(int(ctx_sum) / 1e6 * pin, 4),
+        cost_llm_usd=round(int(in_tok) / 1e6 * pin + int(out_tok) / 1e6 * pout, 4),
     )
 
 

{coderag_ai-0.3.3 → coderag_ai-0.3.4}/src/coderag/api/dashboard.py

@@ -109,6 +109,7 @@ DASHBOARD_HTML = r"""<!doctype html>
   </header>
 
   <section class="kpis" id="kpis"></section>
+  <p style="font-size:12px;color:var(--muted);margin:-14px 0 22px">Dollar figures are estimates at the configured per-million prices (CODERAG_PRICE_INPUT_PER_MTOK / _OUTPUT_PER_MTOK) — not billing data. On a flat-rate Claude plan, savings show up as headroom rather than a refund.</p>
 
   <section class="card">
     <h2>Context vs. saved — recent queries</h2>
@@ -143,6 +144,7 @@ DASHBOARD_HTML = r"""<!doctype html>
 <script>
 const n = x => (x==null? "—" : Number(x).toLocaleString());
 const pct = x => (x==null? "—" : Number(x).toFixed(1) + "%");
+const usd = x => (x==null? "—" : "$" + Number(x).toLocaleString(undefined,{minimumFractionDigits:2, maximumFractionDigits:2}));
 const esc = s => (s||"").replace(/[&<>"]/g, c => ({"&":"&amp;","<":"&lt;",">":"&gt;",'"':"&quot;"}[c]));
 
 function tile(label, val, unit, hero){
@@ -153,7 +155,10 @@ function tile(label, val, unit, hero){
 function renderKpis(m){
   document.getElementById('kpis').innerHTML =
     tile('Saved vs reading files', n(m.total_saved_vs_files), 'tok', true) +
+    tile('Est. $ saved', usd(m.cost_saved_vs_files_usd), '', true) +
     tile('Reduction vs files', pct(m.reduction_vs_files_percent), '', true) +
+    tile('Est. $ context sent', usd(m.cost_context_sent_usd)) +
+    tile('Est. $ LLM spend', usd(m.cost_llm_usd)) +
     tile('Whole-file baseline', n(m.total_baseline_tokens), 'tok') +
     tile('Tokens saved (budgeting)', n(m.total_tokens_saved)) +
     tile('Overall reduction', pct(m.avg_token_reduction_percent)) +

{coderag_ai-0.3.3 → coderag_ai-0.3.4}/src/coderag/api/schemas.py

@@ -158,6 +158,12 @@ class MetricsOut(BaseModel):
     total_baseline_tokens: int
     total_saved_vs_files: int
     reduction_vs_files_percent: float
+    # Indicative USD at the configured per-million prices (estimates, not billing)
+    price_input_per_mtok: float
+    price_output_per_mtok: float
+    cost_saved_vs_files_usd: float
+    cost_context_sent_usd: float
+    cost_llm_usd: float
 
 
 class QueryRow(BaseModel):

{coderag_ai-0.3.3 → coderag_ai-0.3.4}/src/coderag/cli/main.py

@@ -195,9 +195,15 @@ def context(
                       f"{acct.final_prompt_tokens:,}")
             console.print(t)
             if acct.baseline_tokens:
+                price = get_settings().price_input_per_mtok
+                saved_usd = acct.tokens_saved_vs_files / 1e6 * price
                 console.print(
                     f"[green]saved {acct.tokens_saved_vs_files:,} tokens "
-                    f"({acct.reduction_vs_files}%) vs opening those files[/]"
+                    f"({acct.reduction_vs_files}%) vs opening those files[/]  "
+                    f"[dim]≈ ${saved_usd:.4f} at ${price:.2f}/M input tokens; "
+                    f"× 1,000 similar queries ≈ ${saved_usd * 1000:.2f}[/]\n"
+                    f"[dim]Estimate at configured prices, not billing data. On a "
+                    f"flat-rate plan this is headroom, not a refund.[/]"
                 )
             else:
                 console.print("[yellow]No baseline recorded (nothing selected).[/]")

{coderag_ai-0.3.3 → coderag_ai-0.3.4}/src/coderag/core/config.py

@@ -85,6 +85,13 @@ class Settings(BaseSettings):
     llm_max_output_tokens: int = 1024
     llm_timeout_seconds: float = 60.0
 
+    # ---- Cost estimation -------------------------------------------------
+    # USD per million tokens, used only to turn token counts into indicative
+    # dollar figures. Defaults are Claude Opus 4.8 list prices; override for your
+    # model/plan. These are ESTIMATES at configured prices — not billing data.
+    price_input_per_mtok: float = 5.00
+    price_output_per_mtok: float = 25.00
+
     # ---- Analyzers (Phase 10) ------------------------------------------
     max_fix_attempts: int = 2
 

{coderag_ai-0.3.3 → coderag_ai-0.3.4}/tests/test_phase13_measure.py

@@ -77,3 +77,28 @@ def test_mcp_context_includes_measurement(db_session, demo_repo):
     out = coderag_context("retry failed payment", repository="payments")
     assert "fewer input tokens" in out["measurement"]
     assert out["token_accounting"]["baseline_tokens"] > 0
+
+
+def test_metrics_exposes_dollar_estimates(db_session, demo_repo):
+    """Token savings are convertible to indicative USD at configured prices."""
+    from coderag.core.config import get_settings
+
+    run_context(db_session, "retry failed payment", "payments")
+    db_session.commit()
+
+    client = TestClient(app)
+    m = client.get("/metrics").json()
+    s = get_settings()
+    assert m["price_input_per_mtok"] == s.price_input_per_mtok
+    # dollars must be a faithful function of tokens and the configured price
+    expected = round(m["total_saved_vs_files"] / 1e6 * s.price_input_per_mtok, 4)
+    assert m["cost_saved_vs_files_usd"] == expected
+    assert m["cost_context_sent_usd"] >= 0.0
+    assert m["cost_llm_usd"] >= 0.0
+
+
+def test_price_settings_are_configurable():
+    from coderag.core.config import Settings
+
+    s = Settings(price_input_per_mtok=3.0, price_output_per_mtok=15.0)
+    assert s.price_input_per_mtok == 3.0 and s.price_output_per_mtok == 15.0