@winspan/claude-forge 9.2.0 → 9.12.0

package/dist/daemon/templates/agents/hybrid-feature-with-safety.md

@@ -1,104 +0,0 @@
----
-name: hybrid-feature-with-safety
-description: New feature that touches legacy low-coverage code. Use when adding new functionality that requires modifying files in modules with <50% test coverage. Triggers on "新功能 + 改老代码", "add X to legacy module Y", "feature touching hooks/claudemd". For pure new modules use coder; for pure refactor use refactor-specialist.
-tools: Read, Edit, Write, Bash, Glob, Grep
-model: inherit
-color: purple
----
-
-# IMPORTANT: subagent rules
-
-- You are a subagent — you **cannot** spawn other agents (built-in CC restriction)
-- You MUST NOT modify test fixtures to satisfy buggy impl
-- You MUST NOT exceed your declared scope; stop and report if scope creep is needed
-- You MUST run `npm test` (or project equivalent) before declaring done
-- You MUST write a changelog to `docs/implementation/YYYY-MM-DD/HHMM-[task]-changelog.md`
-
-# Role
-
-New feature that touches legacy low-coverage code. You combine the safety-net
-discipline of `refactor-specialist` with the build-new-stuff focus of `coder`.
-You protect the legacy parts you touch from accidental regression while adding
-new capability cleanly alongside.
-
-# When invoked
-
-You should be the agent selected when ALL of these hold:
-
-- User wants to add NEW behavior (feature, capability, endpoint, command)
-- The new behavior touches files in modules with **<50% test coverage**
-  (e.g., `src/hooks/*.sh`, `src/claudemd/`, parts of `src/web/routes/`)
-- Scope spans both "new code" and "modify legacy code"
-
-If new code is wholly isolated → `coder` is fine.
-If nothing new, just restructuring → `refactor-specialist`.
-
-# Workflow
-
-## Step 1: safety-net (legacy touchpoints only)
-
-- Identify exactly which legacy files/functions your feature will modify.
-- For each one, write tests that characterize current behavior (see refactor-specialist Step 1).
-- You do NOT need to test legacy code you only READ; only the parts you'll edit.
-- Run `npm test` — new safety-net tests must pass against unmodified legacy.
-
-## Step 2: analyze
-
-- What does the new feature need from legacy code? Be specific:
-  data, side effects, hooks, configuration?
-- Identify integration points: where exactly will new code call into legacy
-  (or where will legacy call into new)?
-- Note any awkward seams — these are design risks.
-
-## Step 3: design (integration points)
-
-- Brief design note: list integration points and the adapter shape.
-- Prefer ADDING new functions over MODIFYING existing ones (less risk to safety-net).
-- If you must modify a legacy function, justify it in the design note.
-
-## Step 4: implement
-
-- Build the new feature: new files first, then legacy modifications.
-- After each legacy modification, run safety-net tests — must stay green.
-- After each new chunk, run full `npm test` — must stay green.
-
-## Step 5: test
-
-- New feature MUST have its own dedicated tests (the safety-net is only for legacy bits).
-- Cover: happy path, edge cases, error paths, integration with legacy.
-- Aim for the new module to have >=50% coverage from day one.
-
-## Step 6: review
-
-- Diff review: are safety-net tests still passing unchanged? (legacy behavior preserved)
-- Are new tests covering the new code? (feature really tested)
-- Any legacy modifications justified in the design note? (no scope creep)
-
-# Constraints
-
-- Legacy behavior MUST be preserved (safety-net is the contract)
-- New feature MUST have its own tests; do not rely on "well the safety-net still passes"
-  to claim the feature works — that only proves you didn't break old stuff
-- Prefer adding to legacy over modifying legacy; modifying is allowed but documented
-- Same "bugs noticed but not fixed" rule as refactor-specialist
-- **scope-checkpoint**: when this single task's cumulative edits reach 5 distinct
-  files, STOP and report to the main thread (work done so far + remaining plan +
-  any blockers) before continuing — do NOT plow ahead silently; large tasks should
-  be split into batches and re-delegated. (This self-limit counts files within
-  THIS agent's single task; it is independent of the daemon's cross-session
-  `write-multi-file-hard` guardrail — different scopes, different counters.)
-- **verify before done**: before declaring complete you MUST run `npm test` plus
-  any relevant build, and record the PASS/FAIL result explicitly in the changelog.
-  Unverified work may NOT be reported as done.
-
-# Output
-
-- Safety-net test files (NEW)
-- New feature source + tests (NEW)
-- Legacy file modifications (MINIMAL)
-- Changelog with: legacy touchpoints, safety-net test list, new feature design,
-  new test list, integration evidence. **Put a `Decision: <id>` header line at
-  the very top** (e.g. `> Decision: 75344363`) whenever this work traces to an
-  approved decision/spec — this is what lets the Web decision detail page
-  reliably surface the changelog. Omit only if there is no associated id.
-- Final report: "Added feature X (Y LOC new, Z LOC legacy modified); safety-net N tests; new feature M tests"

package/dist/daemon/templates/agents/refactor-specialist.md

@@ -1,98 +0,0 @@
----
-name: refactor-specialist
-description: Behavior-preserving restructure of low-coverage code. Use when user asks to refactor, restructure, clean up, simplify, or extract from a module with <50% test coverage. Triggers on "重构", "refactor", "整理", "拆分". For new features that happen to touch legacy code, use hybrid-feature-with-safety instead.
-tools: Read, Edit, Write, Bash, Glob, Grep
-model: inherit
-color: orange
----
-
-# IMPORTANT: subagent rules
-
-- You are a subagent — you **cannot** spawn other agents (built-in CC restriction)
-- You MUST NOT modify test fixtures to satisfy buggy impl
-- You MUST NOT exceed your declared scope; stop and report if scope creep is needed
-- You MUST run `npm test` (or project equivalent) before declaring done
-- You MUST write a changelog to `docs/implementation/YYYY-MM-DD/HHMM-[task]-changelog.md`
-
-# Role
-
-Behavior-preserving restructure of low-coverage code. Your job is to change the
-SHAPE of code without changing its BEHAVIOR. You characterize current behavior
-with tests first, then move things around, then prove behavior is unchanged.
-
-# When invoked
-
-You should be the agent selected when ALL of these hold:
-
-- User wants to restructure existing code (not add features, not fix bugs)
-- Target module has **<50% test coverage** (you'll need to add safety-net first)
-- Scope is clearly "make it nicer", not "make it do new things"
-
-If the user wants new behavior, even alongside refactoring → use `hybrid-feature-with-safety`.
-If module has good coverage already → standard `coder` is fine.
-
-# Workflow
-
-## Step 1: safety-net (characterize current behavior)
-
-- Read the target module top-to-bottom; understand inputs/outputs.
-- Write tests that pin down current observable behavior:
-  - Pure functions → golden-master tests (input -> exact output)
-  - I/O code → integration tests (set up state, call function, assert state change)
-  - Side-effecty code → spy/mock the boundary, assert call shape
-- Cover the "load-bearing" branches: at least one test per significant code path.
-- Run `npm test` — all new tests must PASS against the unrefactored code.
-- This is the safety net; if you change behavior, these tests will catch it.
-
-## Step 2: design (brief target shape)
-
-- Write a short design note (inline in changelog, or scratch file).
-- Describe: target file layout, what extracts, what merges, what renames.
-- Identify "trickiest move" — start there.
-- Do NOT write a full spec; this is a refactor, the user already approved the direction.
-
-## Step 3: implement (small steps)
-
-- One conceptual move per commit-worthy unit.
-- After each move: run `npm test` — green or revert.
-- If a safety-net test fails, you changed behavior. Revert and rethink.
-- Move from outside-in or inside-out consistently; do not zig-zag.
-
-## Step 4: test (cover the new shape)
-
-- Add tests for any new boundaries you created (extracted functions, new modules).
-- Total coverage should go UP, not just shift around.
-
-## Step 5: review (diff vs original)
-
-- Read your own diff end-to-end.
-- For each chunk, ask: "did I change behavior here?" If yes, revert that chunk.
-- Verify the safety-net tests still pass unchanged (no test edits to accommodate refactor).
-
-# Constraints
-
-- ALL existing behavior MUST be preserved — no "fix bugs along the way"
-- If you find a bug during refactor, note it in changelog as a follow-up;
-  do NOT fix it in the same commit (it muddies the "is this a behavior change?" check)
-- Do NOT add features, do NOT add new public API; only restructure
-- Safety-net tests are immutable once written; if they need editing, your refactor changed behavior
-- **scope-checkpoint**: when this single task's cumulative edits reach 5 distinct
-  files, STOP and report to the main thread (work done so far + remaining plan +
-  any blockers) before continuing — do NOT plow ahead silently; large tasks should
-  be split into batches and re-delegated. (This self-limit counts files within
-  THIS agent's single task; it is independent of the daemon's cross-session
-  `write-multi-file-hard` guardrail — different scopes, different counters.)
-- **verify before done**: before declaring complete you MUST run `npm test` plus
-  any relevant build, and record the PASS/FAIL result explicitly in the changelog.
-  Unverified work may NOT be reported as done.
-
-# Output
-
-- Safety-net test files (NEW, must be added to tree)
-- Refactored source files (behavior unchanged)
-- Changelog with: before/after structure summary, safety-net test list,
-  bugs-noticed-but-not-fixed list, coverage delta. **Put a `Decision: <id>`
-  header line at the very top** (e.g. `> Decision: 75344363`) whenever this work
-  traces to an approved decision/spec — this is what lets the Web decision detail
-  page reliably surface the changelog. Omit only if there is no associated id.
-- Final report: "Refactored X (Y LOC) into Z files; safety-net N tests passing unchanged; coverage X% -> Y%"

package/dist/skills/distill/claude-cli-resolver.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"claude-cli-resolver.d.ts","sourceRoot":"","sources":["../../../src/skills/distill/claude-cli-resolver.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;GAmBG;AAWH,wBAAgB,aAAa,CAAC,IAAI,GAAE;IAAE,KAAK,CAAC,EAAE,OAAO,CAAA;CAAO,GAAG,MAAM,GAAG,IAAI,CAyB3E;AAwDD,4CAA4C;AAC5C,wBAAgB,oBAAoB,IAAI,IAAI,CAG3C"}

package/dist/skills/distill/claude-cli-resolver.js.map

@@ -1 +0,0 @@
-{"version":3,"file":"claude-cli-resolver.js","sourceRoot":"","sources":["../../../src/skills/distill/claude-cli-resolver.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;GAmBG;AAEH,OAAO,EAAE,QAAQ,EAAE,MAAM,oBAAoB,CAAC;AAC9C,OAAO,EAAE,MAAM,SAAS,CAAC;AACzB,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,MAAM,EAAE,MAAM,4BAA4B,CAAC;AAEpD,IAAI,UAAqC,CAAC,CAAC,6BAA6B;AACxE,IAAI,QAAQ,GAAG,CAAC,CAAC;AACjB,MAAM,YAAY,GAAG,CAAC,GAAG,EAAE,GAAG,IAAI,CAAC;AAEnC,MAAM,UAAU,aAAa,CAAC,OAA4B,EAAE;IAC1D,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC;IACvB,IAAI,CAAC,IAAI,CAAC,KAAK,IAAI,UAAU,KAAK,SAAS,IAAI,GAAG,GAAG,QAAQ,GAAG,YAAY,EAAE,CAAC;QAC7E,OAAO,UAAU,CAAC;IACpB,CAAC;IACD,UAAU,GAAG,KAAK,EAAE,CAAC;IACrB,QAAQ,GAAG,GAAG,CAAC;IACf,IAAI,UAAU,EAAE,CAAC;QACf,MAAM,CAAC,IAAI,CAAC,kDAAkD,UAAU,EAAE,CAAC,CAAC;IAC9E,CAAC;SAAM,CAAC;QACN,6EAA6E;QAC7E,uEAAuE;QACvE,4EAA4E;QAC5E,yEAAyE;QACzE,6EAA6E;QAC7E,6EAA6E;QAC7E,4EAA4E;QAC5E,MAAM,CAAC,IAAI,CACT,sFAAsF;YACpF,2BAA2B,OAAO,CAAC,GAAG,CAAC,IAAI,IAAI,EAAE,uCAAuC;YACxF,0FAA0F;YAC1F,0BAA0B,CAC7B,CAAC;IACJ,CAAC;IACD,OAAO,UAAU,CAAC;AACpB,CAAC;AAED;;;;GAIG;AACH,SAAS,gBAAgB;IACvB,MAAM,IAAI,GAAG,OAAO,CAAC,GAAG,CAAC,IAAI,IAAI,EAAE,CAAC;IACpC,MAAM,UAAU,GAAG;QACjB,0BAA0B,EAAE,uCAAuC;QACnE,uBAAuB,EAAE,mCAAmC;QAC5D,GAAG,IAAI,oBAAoB,EAAE,0BAA0B;QACvD,GAAG,IAAI,kBAAkB,EAAE,aAAa;KACzC,CAAC;IACF,6EAA6E;IAC7E,6EAA6E;IAC7E,2EAA2E;IAC3E,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,MAAM,EAAE,UAAU,EAAE,MAAM,CAAC,CAAC;IAC5D,IAAI,CAAC;QACH,MAAM,QAAQ,GAAG,EAAE,CAAC,WAAW,CAAC,OAAO,CAAC,CAAC,IAAI,EAAE,CAAC,OAAO,EAAE,CAAC;QAC1D,KAAK,MAAM,CAAC,IAAI,QAAQ,EAAE,CAAC;YACzB,UAAU,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,OAAO,EAAE,CAAC,EAAE,KAAK,EAAE,QAAQ,CAAC,CAAC,CAAC;QAC1D,CAAC;IACH,CAAC;IAAC,MAAM,CAAC;QACP,oBAAoB;IACtB,CAAC;IACD,+EAA+E;IAC/E,IAAI,IAAI,EAAE,CAAC;QACT,UAAU,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,aAAa,EAAE,KAAK,EAAE,QAAQ,CAAC,CAAC,CAAC;QACjE,UAAU,CAAC,IAAI,CAAC,yCAAyC,CAAC,CAAC;IAC7D,CAAC;IACD,OAAO,UAAU,CAAC;AACpB,CAAC;AAED,SAAS,KAAK;IACZ,uEAAuE;IACvE,IAAI,CAAC;QACH,MAAM,GAAG,GAAG,QAAQ,CAAC,cAAc,EAAE;YACnC,QAAQ,EAAE,OAAO;YACjB,KAAK,EAAE,CAAC,QAAQ,EAAE,MAAM,EAAE,QAAQ,CAAC;SACpC,CAAC,CAAC,IAAI,EAAE,CAAC;QACV,IAAI,GAAG,IAAI,EAAE,CAAC,UAAU,CAAC,GAAG,CAAC;YAAE,OAAO,GAAG,CAAC;IAC5C,CAAC;IAAC,MAAM,CAAC;QACP,cAAc;IAChB,CAAC;IACD,KAAK,MAAM,SAAS,IAAI,gBAAgB,EAAE,EAAE,CAAC;QAC3C,IAAI,CAAC;YACH,IAAI,EAAE,CAAC,UAAU,CAAC,SAAS,CAAC;gBAAE,OAAO,SAAS,CAAC;QACjD,CAAC;QAAC,MAAM,CAAC;YACP,oBAAoB;QACtB,CAAC;IACH,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED,4CAA4C;AAC5C,MAAM,UAAU,oBAAoB;IAClC,UAAU,GAAG,SAAS,CAAC;IACvB,QAAQ,GAAG,CAAC,CAAC;AACf,CAAC"}

package/dist/skills/distilled/distilled-defi-amm-security.md

@@ -1,293 +0,0 @@
----
-name: distilled-defi-amm-security
-version: 0.2.0-distilled
-description: 'Distilled IT-universal skill: defi-amm-security (score 3.5/5)'
-keywords:
-  - defi
-  - amm
-  - security
-rubric_version: 2
-distilled_from:
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - defi-amm-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - django-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - laravel-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - llm-trading-agent-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - perl-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - quarkus-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - security-bounty-hunter/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - security-review/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - security-scan/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - springboot-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - security-review/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - laravel-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - quarkus-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - security-review/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - springboot-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - defi-amm-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - django-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - laravel-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - llm-trading-agent-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - perl-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - security-bounty-hunter/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - security-review/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - security-scan/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - springboot-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - security-review/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - defi-amm-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - django-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - laravel-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - llm-trading-agent-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - perl-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - quarkus-security/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - security-bounty-hunter/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - security-review/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - security-scan/SKILL.md
-  - upstream: https://github.com/affaan-m/ECC.git
-    license: MIT
-    commit: unknown
-    paths:
-      - springboot-security/SKILL.md
-distilled_at: '2026-05-29T08:14:29.853Z'
-distilled_by: claude-code skill-distiller agent
-it_universal_score: 3.5
-it_universal_dims:
-  d1: 0.5
-  d2: 1
-  d3: 0.5
-  d4: 1
-  d5: 0.5
-it_universal_rationale: >-
-  Domain-specific to DeFi/smart-contracts but toolchain-neutral and directly reusable across any AMM
-  protocol audit without project-specific rewrites.
----
-
-# DeFi AMM Smart Contract Security Audit Patterns
-
-## When to use
-- Auditing or reviewing any AMM-style smart contract (swap, mint, burn logic)
-- Detecting economic attack vectors: flash loans, sandwich attacks, price manipulation
-- Adding security assertions to DeFi protocol tests
-- Pre-deployment security review for decentralized exchange contracts
-- Evaluating third-party DeFi integrations your protocol depends on
-
-## Core idea
-AMM contracts face economic attacks that don't exist in traditional software — a
-single transaction can borrow billions, manipulate prices, and repay within the
-same block. The primary defense is the **Check-Effects-Interactions (CEI)**
-pattern combined with trustless price oracles and user-controlled slippage
-guards. Flash loans make any on-chain spot price untrustworthy as a valuation
-oracle; always reference time-weighted or off-chain signed prices. Reentrancy
-guards prevent state corruption when external calls re-enter your contract.
-Access control and event emission provide governance safety and off-chain
-auditability.
-
-## Steps
-
-### 1. Flash Loan Resistance
-- Never trust spot reserve ratios (`reserve1 / reserve0`) as a price oracle
-  inside a single transaction — they are flash-loanable in one block
-- Use TWAP (Time-Weighted Average Price) accumulators sampled over ≥ 30 min,
-  or off-chain signed feeds (e.g., Chainlink, Pyth) for valuation decisions
-- If your function snapshots a price at entry, verify it against a trustworthy
-  oracle *before* any critical state change
-
-### 2. Slippage Protection
-- Every trade/swap function must accept caller-supplied `minAmountOut`
-  (or equivalent) — revert if actual output < `minAmountOut`
-- Add a `deadline` parameter; revert if `block.timestamp > deadline` to
-  prevent transactions from sitting in the mempool and executing at stale rates
-- Never silently accept a worse rate on the user's behalf
-
-### 3. Reentrancy Guards
-- Apply CEI order strictly on every state-changing path:
-  **Check** conditions → **Effect** storage writes → **Interact** externally
-- Add a reentrancy mutex or use a standard `ReentrancyGuard` on swap, mint,
-  burn, and flash-loan entry points
-- Callback tokens (ERC-777, hooks) can re-enter on `transfer` — test this
-  path explicitly
-
-### 4. Price Oracle Integrity
-- **Instantaneous** (`reserve1/reserve0`): manipulable, use only for display
-- **TWAP** (cumulative price accumulators, ≥ 30 min): safer for reference
-- **Off-chain signed feeds**: validate `updatedAt` staleness and deviation
-  thresholds before accepting a price
-- Clearly document which oracle type each function relies on
-
-### 5. Integer Arithmetic Safety
-- Use a compiler version / library with built-in overflow checks, or explicit
-  safe-math throughout
-- In fixed-point math: **multiply before divide** to preserve precision;
-  use adequate decimal precision (18 decimals is standard for ERC-20)
-- For fee calculations in basis points (e.g., 30 bps = 0.3%), verify
-  `amount * fee / 10_000` does not truncate to zero for dust amounts
-
-### 6. Access Control & Governance
-- Separate liquidity-provider roles from protocol-admin roles
-- Privileged operations (fee changes, pause, emergency drain) must be
-  role-gated and ideally time-locked
-- Emit events for every privileged and economically significant state change
-  to enable off-chain monitoring and incident response
-
-## Anti-patterns
-- **Trusting spot reserves as an oracle**: single-block flash loan makes this
-  trivially exploitable
-- **Missing `minAmountOut`**: exposes users to MEV sandwich attacks
-- **State update after external call**: classic reentrancy window
-- **Hardcoded price thresholds**: market conditions change; use relative %
-  deviation guards instead
-- **Single-owner admin key with no timelock**: single point of failure for
-  protocol governance
-- **Silent overflow/underflow**: corrupts pool accounting without revert
-
-## Audit Checklist
-
-- [ ] All valuation-critical price references come from TWAP or external oracle
-- [ ] Flash loan scenario analyzed — protocol is not exploitable within a
-      single block
-- [ ] Every trade function has `minAmountOut` + `deadline` parameters
-- [ ] CEI pattern applied on all state-changing functions with external calls
-- [ ] Reentrancy guard present on swap, mint, burn, and flash entry points
-- [ ] Integer overflow / underflow handled (compiler-enforced or explicit)
-- [ ] Admin vs. LP roles separated; governance actions time-locked
-- [ ] Events emitted for all privileged and economic state changes
-- [ ] Callback-token (ERC-777, hooks) reentrancy explicitly tested in your
-      test suite
-
-## References
-- Upstream: https://github.com/affaan-m/ECC.git

package/dist/skills/keyword-score.d.ts

@@ -1,29 +0,0 @@
-/**
- * Shared keyword-scoring helpers used by SkillRegistry (the only consumer
- * after D R-B; previously also used by the now-deleted semantic matcher).
- *
- * Extracted from registry.ts (matchKeywordFlexible + inner scoring loop) in
- * the semantic-matcher-token-cap spec (2026-05-27, 1249). Behaviour-preserving
- * move — no logic changes.
- */
-import type { Skill } from './registry.js';
-/**
- * Flexible keyword matching:
- * - English single words: requires word boundaries (avoids "test" matching "latest")
- * - Chinese / multi-word: simple substring match
- *
- * Verbatim copy of registry.ts private `matchKeywordFlexible` method.
- */
-export declare function matchKeywordFlexible(text: string, keyword: string): boolean;
-/**
- * Score a single skill against a lower-cased prompt and extra keyword list.
- *
- * Returns a raw integer count (1 per matching keyword in prompt + 1 per
- * matching keyword in extraLower). No threshold or tie-break applied here —
- * those stay in their respective callers.
- *
- * Mirrors the inner scoring loop of registry.ts `matchKeywords` (lines
- * ~273-298) but returns only the numeric score.
- */
-export declare function scoreSkillByKeywords(skill: Skill, promptLower: string, extraLower: string[]): number;
-//# sourceMappingURL=keyword-score.d.ts.map

package/dist/skills/keyword-score.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"keyword-score.d.ts","sourceRoot":"","sources":["../../src/skills/keyword-score.ts"],"names":[],"mappings":"AAAA;;;;;;;GAOG;AAEH,OAAO,KAAK,EAAE,KAAK,EAAE,MAAM,eAAe,CAAC;AAE3C;;;;;;GAMG;AACH,wBAAgB,oBAAoB,CAAC,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,GAAG,OAAO,CAU3E;AAED;;;;;;;;;GASG;AACH,wBAAgB,oBAAoB,CAClC,KAAK,EAAE,KAAK,EACZ,WAAW,EAAE,MAAM,EACnB,UAAU,EAAE,MAAM,EAAE,GACnB,MAAM,CAkBR"}

package/dist/skills/keyword-score.js

@@ -1,54 +0,0 @@
-/**
- * Shared keyword-scoring helpers used by SkillRegistry (the only consumer
- * after D R-B; previously also used by the now-deleted semantic matcher).
- *
- * Extracted from registry.ts (matchKeywordFlexible + inner scoring loop) in
- * the semantic-matcher-token-cap spec (2026-05-27, 1249). Behaviour-preserving
- * move — no logic changes.
- */
-/**
- * Flexible keyword matching:
- * - English single words: requires word boundaries (avoids "test" matching "latest")
- * - Chinese / multi-word: simple substring match
- *
- * Verbatim copy of registry.ts private `matchKeywordFlexible` method.
- */
-export function matchKeywordFlexible(text, keyword) {
-    // For Chinese or multi-word keywords, use simple includes
-    if (/[一-龥]/.test(keyword) || keyword.includes(' ')) {
-        return text.includes(keyword);
-    }
-    // For single English words, require word boundaries to avoid false positives
-    // e.g., "test" won't match "latest"
-    const regex = new RegExp(`\\b${keyword.replace(/[.*+?^${}()|[\]\\]/g, '\\$&')}\\b`);
-    return regex.test(text);
-}
-/**
- * Score a single skill against a lower-cased prompt and extra keyword list.
- *
- * Returns a raw integer count (1 per matching keyword in prompt + 1 per
- * matching keyword in extraLower). No threshold or tie-break applied here —
- * those stay in their respective callers.
- *
- * Mirrors the inner scoring loop of registry.ts `matchKeywords` (lines
- * ~273-298) but returns only the numeric score.
- */
-export function scoreSkillByKeywords(skill, promptLower, extraLower) {
-    if (skill.keywords.length === 0) {
-        // Description-based fallback: used by registry keyword path.
-        // For pre-ranking purposes score stays 0 when no keywords — acceptable.
-        return 0;
-    }
-    let score = 0;
-    for (const kw of skill.keywords) {
-        const kwLower = kw.toLowerCase();
-        if (matchKeywordFlexible(promptLower, kwLower)) {
-            score += 1;
-        }
-        if (extraLower.some(ek => matchKeywordFlexible(ek, kwLower))) {
-            score += 1;
-        }
-    }
-    return score;
-}
-//# sourceMappingURL=keyword-score.js.map

package/dist/skills/keyword-score.js.map

@@ -1 +0,0 @@
-{"version":3,"file":"keyword-score.js","sourceRoot":"","sources":["../../src/skills/keyword-score.ts"],"names":[],"mappings":"AAAA;;;;;;;GAOG;AAIH;;;;;;GAMG;AACH,MAAM,UAAU,oBAAoB,CAAC,IAAY,EAAE,OAAe;IAChE,0DAA0D;IAC1D,IAAI,OAAO,CAAC,IAAI,CAAC,OAAO,CAAC,IAAI,OAAO,CAAC,QAAQ,CAAC,GAAG,CAAC,EAAE,CAAC;QACnD,OAAO,IAAI,CAAC,QAAQ,CAAC,OAAO,CAAC,CAAC;IAChC,CAAC;IAED,6EAA6E;IAC7E,oCAAoC;IACpC,MAAM,KAAK,GAAG,IAAI,MAAM,CAAC,MAAM,OAAO,CAAC,OAAO,CAAC,qBAAqB,EAAE,MAAM,CAAC,KAAK,CAAC,CAAC;IACpF,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AAC1B,CAAC;AAED;;;;;;;;;GASG;AACH,MAAM,UAAU,oBAAoB,CAClC,KAAY,EACZ,WAAmB,EACnB,UAAoB;IAEpB,IAAI,KAAK,CAAC,QAAQ,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QAChC,6DAA6D;QAC7D,wEAAwE;QACxE,OAAO,CAAC,CAAC;IACX,CAAC;IAED,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,KAAK,MAAM,EAAE,IAAI,KAAK,CAAC,QAAQ,EAAE,CAAC;QAChC,MAAM,OAAO,GAAG,EAAE,CAAC,WAAW,EAAE,CAAC;QACjC,IAAI,oBAAoB,CAAC,WAAW,EAAE,OAAO,CAAC,EAAE,CAAC;YAC/C,KAAK,IAAI,CAAC,CAAC;QACb,CAAC;QACD,IAAI,UAAU,CAAC,IAAI,CAAC,EAAE,CAAC,EAAE,CAAC,oBAAoB,CAAC,EAAE,EAAE,OAAO,CAAC,CAAC,EAAE,CAAC;YAC7D,KAAK,IAAI,CAAC,CAAC;QACb,CAAC;IACH,CAAC;IACD,OAAO,KAAK,CAAC;AACf,CAAC"}

package/dist/web/static/assets/AgentContentPage-DkeRNxok.js

@@ -1,2 +0,0 @@
-import{j as e}from"./react-vendor-tkvCrao7.js";import{b as m,L as c}from"./react-router-JVUrkhdd.js";import{u as p}from"./query-S6X1S7K9.js";import{m as s,u as x,T as o,d as y,E as a,Y as u}from"./arco-Bhi3a6Qp.js";import{M as g}from"./MarkdownRenderer-DZmTl-8J.js";import"./vendor-DWgdB1eY.js";import"./syntax-highlighter-BkZfCDsz.js";const j=["name","description","tools","model","version","spawn_agent","tags"];function f(t){if(t==null)return"-";if(Array.isArray(t))return t.join(", ");if(typeof t=="string")return t;if(typeof t=="number"||typeof t=="boolean")return String(t);try{return JSON.stringify(t)}catch{return String(t)}}function I(){const{name:t}=m(),{data:n,isLoading:i,isError:l,error:d}=p({queryKey:["agent-content-page",t],enabled:!!t,queryFn:async()=>{const r=await fetch(`/api/agents/${encodeURIComponent(t)}/content`);if(!r.ok)throw new Error(`HTTP ${r.status}`);return r.json()},staleTime:5*6e4});return e.jsxs("div",{style:{padding:16,display:"flex",flexDirection:"column",gap:12},children:[e.jsxs(s,{children:[e.jsx(s.Item,{children:e.jsx(c,{to:"/context?tab=agents",children:"/context"})}),e.jsx(s.Item,{children:"agent"}),e.jsx(s.Item,{children:e.jsx("span",{style:{fontFamily:"monospace"},children:t})})]}),e.jsxs(x,{bordered:!0,children:[e.jsx("div",{style:{display:"flex",alignItems:"center",gap:12,marginBottom:12},children:e.jsx(o.Title,{heading:5,style:{margin:0,fontFamily:"monospace"},children:t})}),i&&e.jsx(y,{animation:!0,text:{rows:10}}),l&&e.jsx(a,{description:e.jsxs(o.Text,{type:"error",children:["加载失败: ",String(d)]})}),n&&!n.exists&&e.jsx(a,{description:e.jsxs("span",{children:["未找到 ",e.jsxs("code",{style:{fontFamily:"monospace"},children:[".claude/agents/",t,".md"]}),"。",e.jsx("br",{}),"这通常是 Claude Code 内置的虚拟 subagent_type（如 ",e.jsx("code",{children:"general-purpose"}),"），无对应的 markdown 文件。"]})}),n&&n.exists&&e.jsxs(e.Fragment,{children:[e.jsx(u,{colon:" :",size:"small",column:2,title:e.jsx(o.Text,{style:{fontSize:13},children:"Frontmatter"}),data:j.map(r=>({label:r,value:f(n.frontmatter[r])})),style:{marginBottom:16}}),e.jsxs(o.Text,{type:"secondary",style:{fontSize:12},children:["path: ",n.path," · ",n.bytes," bytes"]}),e.jsx("div",{style:{marginTop:12},children:e.jsx(g,{content:n.body})})]})]})]})}export{I as default};
-//# sourceMappingURL=AgentContentPage-DkeRNxok.js.map

package/dist/web/static/assets/AgentContentPage-DkeRNxok.js.map

@@ -1 +0,0 @@
-{"version":3,"file":"AgentContentPage-DkeRNxok.js","sources":["../../src/pages/AgentContentPage.tsx"],"sourcesContent":["/**\n * /context/agent/:name — full-page view of an agent's .claude/agents/<name>.md.\n *\n * Standalone fallback for long agent definitions where the 760px drawer is\n * cramped. Linked from AgentContentDrawer's \"Open in full page\" anchor.\n *\n * Breadcrumb: /context → /context/agent/:name\n */\nimport { useParams, Link } from 'react-router-dom'\nimport { useQuery } from '@tanstack/react-query'\nimport Breadcrumb from '@arco-design/web-react/es/Breadcrumb'\nimport Card from '@arco-design/web-react/es/Card'\nimport Descriptions from '@arco-design/web-react/es/Descriptions'\nimport Empty from '@arco-design/web-react/es/Empty'\nimport Skeleton from '@arco-design/web-react/es/Skeleton'\nimport Typography from '@arco-design/web-react/es/Typography'\nimport MarkdownRenderer from '../components/MarkdownRenderer'\n\ninterface AgentContentResponse {\n  name: string\n  exists: boolean\n  path: string | null\n  bytes: number\n  body: string\n  frontmatter: Record<string, unknown>\n  tools: string | null\n  model: string | null\n}\n\nconst WHITELIST: ReadonlyArray<string> = [\n  'name',\n  'description',\n  'tools',\n  'model',\n  'version',\n  'spawn_agent',\n  'tags',\n]\n\nfunction renderFmValue(v: unknown): string {\n  if (v == null) return '-'\n  if (Array.isArray(v)) return v.join(', ')\n  if (typeof v === 'string') return v\n  if (typeof v === 'number' || typeof v === 'boolean') return String(v)\n  try {\n    return JSON.stringify(v)\n  } catch {\n    return String(v)\n  }\n}\n\nexport default function AgentContentPage() {\n  const { name } = useParams<{ name: string }>()\n  const { data, isLoading, isError, error } = useQuery<AgentContentResponse>({\n    queryKey: ['agent-content-page', name],\n    enabled: !!name,\n    queryFn: async () => {\n      const r = await fetch(`/api/agents/${encodeURIComponent(name!)}/content`)\n      if (!r.ok) throw new Error(`HTTP ${r.status}`)\n      return r.json()\n    },\n    staleTime: 5 * 60_000,\n  })\n\n  return (\n    <div style={{ padding: 16, display: 'flex', flexDirection: 'column', gap: 12 }}>\n      <Breadcrumb>\n        <Breadcrumb.Item>\n          <Link to=\"/context?tab=agents\">/context</Link>\n        </Breadcrumb.Item>\n        <Breadcrumb.Item>agent</Breadcrumb.Item>\n        <Breadcrumb.Item>\n          <span style={{ fontFamily: 'monospace' }}>{name}</span>\n        </Breadcrumb.Item>\n      </Breadcrumb>\n\n      <Card bordered>\n        <div style={{ display: 'flex', alignItems: 'center', gap: 12, marginBottom: 12 }}>\n          <Typography.Title heading={5} style={{ margin: 0, fontFamily: 'monospace' }}>{name}</Typography.Title>\n        </div>\n\n        {isLoading && <Skeleton animation text={{ rows: 10 }} />}\n        {isError && (\n          <Empty description={<Typography.Text type=\"error\">加载失败: {String(error)}</Typography.Text>} />\n        )}\n        {data && !data.exists && (\n          <Empty\n            description={\n              <span>\n                未找到 <code style={{ fontFamily: 'monospace' }}>.claude/agents/{name}.md</code>。\n                <br />\n                这通常是 Claude Code 内置的虚拟 subagent_type（如 <code>general-purpose</code>），无对应的 markdown 文件。\n              </span>\n            }\n          />\n        )}\n        {data && data.exists && (\n          <>\n            <Descriptions\n              colon=\" :\"\n              size=\"small\"\n              column={2}\n              title={<Typography.Text style={{ fontSize: 13 }}>Frontmatter</Typography.Text>}\n              data={WHITELIST.map((key) => ({\n                label: key,\n                value: renderFmValue(data.frontmatter[key]),\n              }))}\n              style={{ marginBottom: 16 }}\n            />\n            <Typography.Text type=\"secondary\" style={{ fontSize: 12 }}>\n              path: {data.path} · {data.bytes} bytes\n            </Typography.Text>\n            <div style={{ marginTop: 12 }}>\n              <MarkdownRenderer content={data.body} />\n            </div>\n          </>\n        )}\n      </Card>\n    </div>\n  )\n}\n"],"names":["WHITELIST","renderFmValue","v","AgentContentPage","name","useParams","data","isLoading","isError","error","useQuery","jsxs","Breadcrumb","jsx","Link","Card","Typography","Skeleton","Empty","Fragment","Descriptions","key","MarkdownRenderer"],"mappings":"gVA6BA,MAAMA,EAAmC,CACvC,OACA,cACA,QACA,QACA,UACA,cACA,MACF,EAEA,SAASC,EAAcC,EAAoB,CACzC,GAAIA,GAAK,KAAM,MAAO,IACtB,GAAI,MAAM,QAAQA,CAAC,EAAG,OAAOA,EAAE,KAAK,IAAI,EACxC,GAAI,OAAOA,GAAM,SAAU,OAAOA,EAClC,GAAI,OAAOA,GAAM,UAAY,OAAOA,GAAM,UAAW,OAAO,OAAOA,CAAC,EACpE,GAAI,CACF,OAAO,KAAK,UAAUA,CAAC,CACzB,MAAQ,CACN,OAAO,OAAOA,CAAC,CACjB,CACF,CAEA,SAAwBC,GAAmB,CACzC,KAAM,CAAE,KAAAC,CAAA,EAASC,EAAA,EACX,CAAE,KAAAC,EAAM,UAAAC,EAAW,QAAAC,EAAS,MAAAC,CAAA,EAAUC,EAA+B,CACzE,SAAU,CAAC,qBAAsBN,CAAI,EACrC,QAAS,CAAC,CAACA,EACX,QAAS,SAAY,CACnB,MAAM,EAAI,MAAM,MAAM,eAAe,mBAAmBA,CAAK,CAAC,UAAU,EACxE,GAAI,CAAC,EAAE,GAAI,MAAM,IAAI,MAAM,QAAQ,EAAE,MAAM,EAAE,EAC7C,OAAO,EAAE,KAAA,CACX,EACA,UAAW,EAAI,GAAA,CAChB,EAED,OACEO,EAAAA,KAAC,MAAA,CAAI,MAAO,CAAE,QAAS,GAAI,QAAS,OAAQ,cAAe,SAAU,IAAK,EAAA,EACxE,SAAA,CAAAA,OAACC,EAAA,CACC,SAAA,CAAAC,EAAAA,IAACD,EAAW,KAAX,CACC,SAAAC,EAAAA,IAACC,GAAK,GAAG,sBAAsB,oBAAQ,CAAA,CACzC,EACAD,EAAAA,IAACD,EAAW,KAAX,CAAgB,SAAA,OAAA,CAAK,EACtBC,EAAAA,IAACD,EAAW,KAAX,CACC,SAAAC,EAAAA,IAAC,OAAA,CAAK,MAAO,CAAE,WAAY,WAAA,EAAgB,SAAAT,CAAA,CAAK,CAAA,CAClD,CAAA,EACF,EAEAO,EAAAA,KAACI,EAAA,CAAK,SAAQ,GACZ,SAAA,CAAAF,EAAAA,IAAC,MAAA,CAAI,MAAO,CAAE,QAAS,OAAQ,WAAY,SAAU,IAAK,GAAI,aAAc,EAAA,EAC1E,SAAAA,EAAAA,IAACG,EAAW,MAAX,CAAiB,QAAS,EAAG,MAAO,CAAE,OAAQ,EAAG,WAAY,WAAA,EAAgB,SAAAZ,CAAA,CAAK,CAAA,CACrF,EAECG,SAAcU,EAAA,CAAS,UAAS,GAAC,KAAM,CAAE,KAAM,EAAA,EAAM,EACrDT,SACEU,EAAA,CAAM,mBAAcF,EAAW,KAAX,CAAgB,KAAK,QAAQ,SAAA,CAAA,SAAO,OAAOP,CAAK,CAAA,CAAA,CAAE,CAAA,CAAoB,EAE5FH,GAAQ,CAACA,EAAK,QACbO,EAAAA,IAACK,EAAA,CACC,mBACG,OAAA,CAAK,SAAA,CAAA,cACC,OAAA,CAAK,MAAO,CAAE,WAAY,aAAe,SAAA,CAAA,kBAAgBd,EAAK,KAAA,EAAG,EAAO,UAC5E,KAAA,EAAG,EAAE,0CACiCS,EAAAA,IAAC,QAAK,SAAA,iBAAA,CAAe,EAAO,qBAAA,CAAA,CACrE,CAAA,CAAA,EAILP,GAAQA,EAAK,QACZK,EAAAA,KAAAQ,EAAAA,SAAA,CACE,SAAA,CAAAN,EAAAA,IAACO,EAAA,CACC,MAAM,KACN,KAAK,QACL,OAAQ,EACR,MAAOP,EAAAA,IAACG,EAAW,KAAX,CAAgB,MAAO,CAAE,SAAU,EAAA,EAAM,SAAA,aAAA,CAAW,EAC5D,KAAMhB,EAAU,IAAKqB,IAAS,CAC5B,MAAOA,EACP,MAAOpB,EAAcK,EAAK,YAAYe,CAAG,CAAC,CAAA,EAC1C,EACF,MAAO,CAAE,aAAc,EAAA,CAAG,CAAA,EAE5BV,EAAAA,KAACK,EAAW,KAAX,CAAgB,KAAK,YAAY,MAAO,CAAE,SAAU,EAAA,EAAM,SAAA,CAAA,SAClDV,EAAK,KAAK,MAAIA,EAAK,MAAM,QAAA,EAClC,EACAO,EAAAA,IAAC,MAAA,CAAI,MAAO,CAAE,UAAW,EAAA,EACvB,SAAAA,EAAAA,IAACS,EAAA,CAAiB,QAAShB,EAAK,IAAA,CAAM,CAAA,CACxC,CAAA,CAAA,CACF,CAAA,CAAA,CAEJ,CAAA,EACF,CAEJ"}

package/dist/web/static/assets/AgentDelegationTable-ByBa0x1l.js

@@ -1,2 +0,0 @@
-import{j as e,r as x}from"./react-vendor-tkvCrao7.js";import{u as y}from"./query-S6X1S7K9.js";import{U as h,Y as g,d as f,E as l,T as i,ai as j,X as b,b as u}from"./arco-Bhi3a6Qp.js";import{M as w}from"./MarkdownRenderer-DZmTl-8J.js";import"./vendor-DWgdB1eY.js";import"./syntax-highlighter-BkZfCDsz.js";const T=["name","description","tools","model","version","spawn_agent","tags"];function S(n){if(n==null)return"-";if(Array.isArray(n))return n.join(", ");if(typeof n=="string")return n;if(typeof n=="number"||typeof n=="boolean")return String(n);try{return JSON.stringify(n)}catch{return String(n)}}function z({open:n,agentName:a,stats:o,onClose:c}){const{data:r,isLoading:p,isError:d,error:m}=y({queryKey:["agent-content",a],enabled:!!a&&n,queryFn:async()=>{const s=await fetch(`/api/agents/${encodeURIComponent(a)}/content`);if(!s.ok)throw new Error(`HTTP ${s.status}`);return s.json()},staleTime:3e5}),t=e.jsxs("div",{style:{display:"flex",alignItems:"center",gap:12,flexWrap:"wrap"},children:[e.jsx("span",{style:{fontFamily:"monospace",fontSize:13},children:a??"-"}),a&&e.jsxs("a",{href:`/context/agent/${encodeURIComponent(a)}`,target:"_blank",rel:"noopener noreferrer",style:{marginLeft:"auto",fontSize:12,display:"inline-flex",alignItems:"center",gap:4},children:[e.jsx(j,{})," Open in full page"]})]});return e.jsxs(h,{visible:n,onCancel:c,onOk:c,title:t,width:760,footer:null,unmountOnExit:!0,children:[o&&e.jsx(g,{colon:" :",size:"small",column:3,data:[{label:"delegations (7d)",value:o.count},{label:"rate",value:o.rate},{label:"subagent_type",value:o.agent}],style:{marginBottom:12}}),p&&e.jsx(f,{animation:!0,text:{rows:6}}),d&&e.jsx(l,{description:e.jsxs(i.Text,{type:"error",children:["加载失败: ",String(m)]})}),r&&!r.exists&&r.source==="builtin"&&e.jsx(l,{description:e.jsxs("span",{children:[e.jsx("code",{style:{fontFamily:"monospace"},children:a})," 是 Claude Code 内置的虚拟 subagent_type，无项目级 markdown 文件。",e.jsx("br",{}),r.message??"Claude Code built-in agent; no project-level details available"]})}),r&&!r.exists&&r.source!=="builtin"&&e.jsx(l,{description:e.jsxs("span",{children:["未找到 ",e.jsxs("code",{style:{fontFamily:"monospace"},children:[".claude/agents/",a,".md"]}),"。",e.jsx("br",{}),"若这是新建 agent，请在 TUI 重启（Cmd+R）后再触发。"]})}),r&&r.exists&&e.jsxs(e.Fragment,{children:[e.jsx(g,{colon:" :",size:"mini",column:2,title:e.jsx(i.Text,{style:{fontSize:13},children:"Frontmatter"}),data:T.map(s=>({label:s,value:S(r.frontmatter[s])})),style:{marginBottom:16}}),e.jsx(i.Text,{style:{fontSize:13,fontWeight:600},children:"Body"}),e.jsx("div",{style:{marginTop:8,maxHeight:"calc(100vh - 360px)",overflow:"auto",border:"1px solid var(--color-border-2)",borderRadius:4,padding:12},children:e.jsx(w,{content:r.body})})]})]})}function R(){const[n,a]=x.useState(null),{data:o,isLoading:c,isError:r,error:p}=y({queryKey:["ctx-agent-hit-rate"],queryFn:async()=>{const t=await fetch("/api/rules/hit-rate?days=7");if(!t.ok)throw new Error(`HTTP ${t.status}`);return t.json()},staleTime:6e4}),d=x.useMemo(()=>o?o.agents.map(t=>({...t})).sort((t,s)=>s.count-t.count):[],[o]);if(r)return e.jsx(l,{description:e.jsxs(i.Text,{type:"error",children:["加载失败: ",String(p)]})});if(c||!o)return e.jsx(f,{animation:!0,text:{rows:8}});const m=[{title:"agent (subagent_type)",dataIndex:"agent",render:(t,s)=>e.jsxs("span",{style:{display:"inline-flex",alignItems:"center",gap:6},children:[e.jsx("span",{style:{fontFamily:"monospace",fontSize:12},children:t}),s.source==="builtin"&&e.jsx(u,{color:"gray",size:"small",children:"built-in"})]})},{title:"delegations (7d)",dataIndex:"count",width:160,sorter:(t,s)=>t.count-s.count,render:t=>e.jsx("span",{style:{fontFamily:"monospace",fontWeight:600},children:t})},{title:"rate (vs 全部 prompt)",dataIndex:"rate",width:160,render:t=>e.jsx(i.Text,{style:{fontSize:12},children:t})},{title:"热度",dataIndex:"count",key:"tier",width:100,render:t=>t>=50?e.jsx(u,{color:"green",size:"small",children:"heavy"}):t>=5?e.jsx(u,{color:"arcoblue",size:"small",children:"used"}):e.jsx(u,{color:"orange",size:"small",children:"rare"})}];return e.jsxs("div",{children:[e.jsx("div",{style:{marginBottom:12},children:e.jsxs(i.Text,{style:{fontSize:13,color:"var(--color-text-3)"},children:["近 7d · 总 delegation ",o.summary.agentDelegations," / ",o.summary.totalPrompts," prompts · agent rate ",o.summary.agentRate," · 共 ",d.length," 个 agent"]})}),e.jsx(b,{rowKey:"agent",data:d,columns:m,pagination:{pageSize:20,showTotal:!0},size:"small",border:{wrapper:!0,cell:!0},onRow:t=>({onClick:()=>a(t),style:{cursor:"pointer"}}),noDataElement:e.jsx(l,{description:"无 agent 委托记录"})}),e.jsx(z,{open:n!==null,agentName:(n==null?void 0:n.agent)??null,stats:n,onClose:()=>a(null)})]})}export{R as default};
-//# sourceMappingURL=AgentDelegationTable-ByBa0x1l.js.map