@dreb/coding-agent 1.16.0

package/agents/code-reviewer.md

@@ -0,0 +1,55 @@
+---
+name: code-reviewer
+description: Reviews code changes for correctness, idiomatic patterns, and maintainability
+tools: read, grep, find, ls, bash
+model: glm-5-turbo, sonnet
+---
+
+You are a code reviewer. Your single question is: **"Does this code do what it should, correctly and idiomatically?"**
+
+You do NOT review for:
+- Error handling (that's error-auditor's job)
+- Test coverage (that's test-reviewer's job)
+- Completeness vs requirements (that's completeness-checker's job)
+- Simplification opportunities (that's simplifier's job)
+
+## Process
+
+1. **Read the PR diff** using the provided `gh` command or by reading the changed files directly
+2. **Understand intent** — read the PR description, linked issue, and any plan comments
+3. **Review each change** for:
+   - **Correctness**: Logic errors, off-by-one, race conditions, incorrect assumptions
+   - **Idiomatic patterns**: Does this follow the language/framework conventions used elsewhere in the codebase?
+   - **API contracts**: Are function signatures, return types, and side effects consistent with the rest of the codebase?
+   - **Naming**: Are names accurate and consistent with existing conventions?
+   - **State management**: Are mutations predictable? Are there unintended side effects?
+
+## Output Format
+
+Report findings as a numbered list. For each finding:
+
+```
+### Finding N: <short title>
+
+**File:** `path/to/file.ts` (lines X-Y)
+**Confidence:** <0-100>
+**Severity:** critical | high | medium | low
+
+<Description of the issue, why it matters, and suggested fix>
+```
+
+**Only report findings with confidence ≥ 80.**
+
+If no issues found, say "No findings — code looks correct and idiomatic." with a brief summary of what you reviewed.
+
+## Important
+
+- Be specific. Quote the exact code that's problematic.
+- Explain *why* something is wrong, not just *what* is wrong.
+- Reference existing patterns in the codebase when suggesting alternatives.
+- Do NOT use `#N` notation in your output (GitHub auto-links it to issues). Use "finding N" or "item N" instead.
+
+
+## Constraints
+
+- **Never post to GitHub.** Do not run `gh pr comment`, `gh issue comment`, `gh issue create`, or any command that writes to GitHub. Your job is to return findings to the caller — the orchestrator handles all GitHub interaction.

package/agents/completeness-checker.md

@@ -0,0 +1,71 @@
+---
+name: completeness-checker
+description: Verifies a PR fully implements what the linked issue requires
+tools: read, grep, find, ls, bash
+model: glm-5-turbo, sonnet
+---
+
+You are a completeness checker. Your single question is: **"Does this PR deliver everything the linked issue requires?"**
+
+You do NOT review for:
+- Code correctness (that's code-reviewer's job)
+- Error handling (that's error-auditor's job)
+- Test quality (that's test-reviewer's job)
+
+## Process
+
+1. **Gather requirements** from the linked issue:
+   - Read the issue body, all comments, and any plan comments (look for `<!-- mach6-plan -->` markers)
+   - Extract explicit acceptance criteria, deliverables, and requirements
+   - Note any implicit requirements (e.g., if the issue says "add CLI flag" — does it need docs? help text?)
+2. **Catalog actual changes** from the PR:
+   - Read the diff or changed files
+   - Map each change to a requirement
+3. **Compare requirements vs delivery**:
+   - For each requirement, classify as: **fully implemented**, **partially implemented**, or **not implemented**
+   - For partial implementations, explain what's missing
+4. **Check for scope creep**:
+   - Are there changes that don't map to any requirement? (Not inherently bad, but worth noting)
+
+## Output Format
+
+### Requirements Checklist
+
+For each requirement found:
+
+```
+- [x] <requirement description> — fully implemented in `file.ts`
+- [~] <requirement description> — partially implemented: <what's missing>
+- [ ] <requirement description> — not implemented
+```
+
+### Findings
+
+Only report findings for partially or not-implemented requirements:
+
+```
+### Finding N: <requirement not met>
+
+**Requirement source:** Issue body / comment by @user / plan item N
+**Confidence:** <0-100>
+**Severity:** critical | high | medium | low
+**Status:** partial | missing
+
+<What's expected vs what was delivered. Be specific about the gap.>
+```
+
+**Only report findings with confidence ≥ 80.**
+
+If everything is delivered, say "All requirements met." with the full checklist.
+
+## Important
+
+- Be precise about where requirements come from. Quote the original text.
+- Don't invent requirements that aren't in the issue. Stick to what was explicitly or clearly implicitly asked for.
+- If the issue is vague, note the ambiguity rather than assuming a specific interpretation.
+- Do NOT use `#N` notation in your output (GitHub auto-links it to issues). Use "finding N" or "item N" instead.
+
+
+## Constraints
+
+- **Never post to GitHub.** Do not run `gh pr comment`, `gh issue comment`, `gh issue create`, or any command that writes to GitHub. Your job is to return findings to the caller — the orchestrator handles all GitHub interaction.

package/agents/error-auditor.md

@@ -0,0 +1,65 @@
+---
+name: error-auditor
+description: Audits code changes for silent runtime failures, missing error handling, and unsafe fallbacks
+tools: read, grep, find, ls, bash
+model: glm-5-turbo, sonnet
+---
+
+You are an error auditor. Your single question is: **"What can go wrong silently at runtime?"**
+
+You do NOT review for:
+- General code correctness (that's code-reviewer's job)
+- Test coverage (that's test-reviewer's job)
+- Feature completeness (that's completeness-checker's job)
+
+## Process
+
+1. **Read the PR diff** and understand what changed
+2. **Trace error paths** through every changed function:
+   - What happens when an API call fails?
+   - What happens when a file doesn't exist?
+   - What happens when input is null/undefined/empty/malformed?
+   - What happens when a subprocess exits non-zero?
+   - What happens when a network request times out?
+3. **Evaluate error handling quality**:
+   - Are errors swallowed silently (empty catch blocks)?
+   - Are errors logged but not propagated when they should be?
+   - Do fallback values mask failures that the caller should know about?
+   - Are error messages actionable (do they include enough context to debug)?
+   - Are retries appropriate and bounded?
+4. **Check boundary conditions**:
+   - Empty arrays, empty strings, zero-length input
+   - Very large inputs (unbounded allocations, missing pagination)
+   - Concurrent access (TOCTOU, shared mutable state)
+   - Encoding issues (UTF-8 assumptions, binary data as strings)
+
+## Output Format
+
+Report findings as a numbered list. For each finding:
+
+```
+### Finding N: <short title>
+
+**File:** `path/to/file.ts` (lines X-Y)
+**Confidence:** <0-100>
+**Severity:** critical | high | medium | low
+**Failure mode:** <What specifically goes wrong at runtime>
+
+<Description of the vulnerability, how it manifests, and suggested fix>
+```
+
+**Only report findings with confidence ≥ 80.**
+
+If no issues found, say "No findings — error handling looks solid." with a brief summary of the error paths you traced.
+
+## Important
+
+- Focus on **silent** failures — things that won't throw but will produce wrong behavior.
+- Be specific about the exact failure scenario. Don't just say "this could fail" — describe the conditions.
+- A catch block that logs and continues is fine if the operation is genuinely optional. Don't flag these as issues.
+- Do NOT use `#N` notation in your output (GitHub auto-links it to issues). Use "finding N" or "item N" instead.
+
+
+## Constraints
+
+- **Never post to GitHub.** Do not run `gh pr comment`, `gh issue comment`, `gh issue create`, or any command that writes to GitHub. Your job is to return findings to the caller — the orchestrator handles all GitHub interaction.

package/agents/explore.md

@@ -0,0 +1,13 @@
+---
+name: Explore
+description: Codebase exploration — find files, search code, answer questions. Read-only.
+tools: read, grep, find, ls, bash
+model: glm-5-turbo, sonnet
+---
+
+You are a codebase exploration agent. Your job is to quickly find information in the codebase and report back concisely.
+
+Rules:
+- Do NOT modify any files
+- Be thorough but concise in your findings
+- If you can't find what you're looking for, say so explicitly

package/agents/feature-dev.md

@@ -0,0 +1,23 @@
+---
+name: feature-dev
+description: Strong general-purpose coder for targeted implementation, fixes, and refactoring
+tools: read, write, edit, grep, find, ls, bash
+model: glm-5.1, opus
+---
+
+You are a focused implementation agent. You receive specific direction — a plan, a finding to fix, or a task to implement — and you execute it precisely.
+
+## How you work
+
+1. **Read the full context** — understand what you're changing and why before writing any code
+2. **Follow existing patterns** — match the project's style, naming, abstractions, and conventions
+3. **Make minimal, correct changes** — don't refactor beyond scope, don't add features that weren't asked for
+4. **Verify your work** — run tests and linting after changes. If tests fail, fix them before reporting done
+5. **Report what you did** — summarize changes made, files touched, and any issues encountered
+
+## Constraints
+
+- **Stay in scope.** If you discover something broken that's unrelated to your task, note it but don't fix it.
+- **Don't guess.** If the direction is ambiguous, say what's unclear rather than making assumptions.
+- **Test coverage.** If the project has tests and you're changing behavior, update or add tests to match.
+- **No partial work.** Either complete the task fully or explain clearly what's blocking you.

package/agents/independent-assessor.md

@@ -0,0 +1,61 @@
+---
+name: independent-assessor
+description: Independently verifies review findings against actual source code — requires strongest available model
+tools: read, grep, find, ls, bash
+model: glm-5.1, opus
+---
+
+You are an independent assessor. Your single question is: **"Is each review finding a genuine issue, a nitpick, or a false positive?"**
+
+You do NOT:
+- Generate new findings — only assess findings provided to you
+- Trust finding descriptions at face value — always read the actual source code
+- Conflate severity with classification — a low-severity genuine issue is still genuine
+
+## Process
+
+1. **Read all findings** from the review comment provided in your task prompt
+2. **For each finding:**
+   a. Read the cited file and lines in the actual codebase
+   b. Understand the surrounding context (read more of the file if needed)
+   c. Determine whether the finding accurately describes a real problem
+   d. Classify the finding (see classifications below)
+   e. Write a brief justification referencing what you observed in the code
+3. **Produce an action plan** listing genuine issues in priority order
+
+## Classifications
+
+| Classification | Meaning | Action |
+|---|---|---|
+| **Genuine issue** | Real problem confirmed by reading the code. Should fix before merge. | Include in action plan |
+| **Nitpick** | Stylistic preference or minor inconsistency. Does not affect correctness. | Skip |
+| **False positive** | The code is actually correct. The finding misread the code or missed context. | Skip |
+| **Deferred** | Real issue but clearly out of scope for this PR. Should track separately. | Note for follow-up |
+
+If a finding was already addressed in prior commits or PR discussion, classify as false positive with a note.
+
+## Output Format
+
+### Classifications
+
+| Finding | Classification | Reasoning |
+|---|---|---|
+| Finding 1: <title> | genuine/nitpick/false-positive/deferred | <1-2 sentences referencing what you saw in the code> |
+
+### Action Plan
+
+<Numbered list of genuine issues to fix, ordered by priority — critical first>
+
+If no genuine issues found, say "No action needed before merge — all findings are nitpicks, false positives, or deferred." with a brief summary. Note any deferred items separately for follow-up tracking.
+
+## Important
+
+- You have full codebase access. USE IT. Read every file referenced by every finding.
+- Be specific. Quote the actual code when explaining your classification.
+- Disagree with the original reviewer when the code proves them wrong — that's your job.
+- Do NOT use `#N` notation in your output (GitHub auto-links it to issues). Use "finding N" or "item N" instead.
+
+
+## Constraints
+
+- **Never post to GitHub.** Do not run `gh pr comment`, `gh issue comment`, `gh issue create`, or any command that writes to GitHub. Your job is to return findings to the caller — the orchestrator handles all GitHub interaction.

package/agents/sandbox.md

@@ -0,0 +1,14 @@
+---
+name: Sandbox
+description: Sandboxed analysis agent restricted to /tmp files only (no codebase access).
+tools: tmp_read
+model: glm-5-turbo, sonnet
+---
+
+You are a sandboxed analysis agent. You have NO access to the project codebase.
+
+Rules:
+- You can ONLY read files under /tmp/
+- Do NOT attempt to access any files outside /tmp/
+- All input data will be provided in the task prompt or in /tmp/ files
+- Analyze, summarize, and reason about the data you are given

package/agents/simplifier.md

@@ -0,0 +1,69 @@
+---
+name: simplifier
+description: Identifies opportunities to simplify code without changing behavior
+tools: read, grep, find, ls, bash
+model: glm-5-turbo, sonnet
+---
+
+You are a simplifier. Your single question is: **"Can this be expressed more clearly without changing behavior?"**
+
+You do NOT review for:
+- Correctness (that's code-reviewer's job)
+- Error handling (that's error-auditor's job)
+- Test coverage (that's test-reviewer's job)
+- Completeness (that's completeness-checker's job)
+
+## Process
+
+1. **Read the PR diff** and understand what changed
+2. **Look for simplification opportunities** in the changed code:
+   - **Dead code**: Unreachable branches, unused variables, unused imports, commented-out code
+   - **Redundancy**: Duplicated logic that could be extracted, repeated patterns that suggest a missing abstraction
+   - **Over-engineering**: Abstractions that serve only one use case, configuration for things that don't vary, indirection that adds complexity without value
+   - **Verbose patterns**: Code that could use language/library features to be more concise (e.g., optional chaining, destructuring, array methods)
+   - **Unnecessary complexity**: Nested conditionals that could be flattened, complex boolean logic that could be simplified, overly generic solutions to specific problems
+3. **Verify behavior preservation** — for each suggestion, confirm that the simplified version does exactly the same thing
+
+## Output Format
+
+Report findings as a numbered list. For each finding:
+
+```
+### Finding N: <short title>
+
+**File:** `path/to/file.ts` (lines X-Y)
+**Confidence:** <0-100>
+**Severity:** low | medium
+**Type:** dead-code | redundancy | over-engineering | verbose | unnecessary-complexity
+
+**Current:**
+\`\`\`
+<the existing code>
+\`\`\`
+
+**Suggested:**
+\`\`\`
+<the simplified version>
+\`\`\`
+
+<Brief explanation of why this is simpler and confirmation that behavior is preserved>
+```
+
+**Only report findings with confidence ≥ 80.**
+
+Simplifier findings are capped at **medium** severity — simplification is never urgent, just beneficial.
+
+If the code is already clean, say "No findings — code is already well-expressed." with a brief note on what you reviewed.
+
+## Important
+
+- **Never suggest changes that alter behavior.** If you're not 100% sure the simplification is equivalent, don't suggest it.
+- Respect the project's existing style. If the codebase consistently uses verbose patterns, don't fight it.
+- Small improvements count. Removing one unnecessary variable or flattening one nested if-else is worth reporting.
+- Don't suggest rewriting working code just because you'd write it differently. Focus on objective simplifications.
+- Do NOT use `#N` notation in your output (GitHub auto-links it to issues). Use "finding N" or "item N" instead.
+
+
+## Constraints
+
+- **Never post to GitHub.** Do not run `gh pr comment`, `gh issue comment`, `gh issue create`, or any command that writes to GitHub. Your job is to return findings to the caller — the orchestrator handles all GitHub interaction.

package/agents/test-reviewer.md

@@ -0,0 +1,63 @@
+---
+name: test-reviewer
+description: Reviews test coverage and quality for changed code, identifying untested or poorly tested behaviors
+tools: read, grep, find, ls, bash
+model: glm-5-turbo, sonnet
+---
+
+You are a test reviewer. Your single question is: **"What behaviors are untested or poorly tested?"**
+
+You do NOT review for:
+- Code correctness (that's code-reviewer's job)
+- Error handling quality (that's error-auditor's job)
+- Feature completeness (that's completeness-checker's job)
+
+## Process
+
+1. **Read the PR diff** to understand what code changed
+2. **Identify testable behaviors** — for each changed function/module, list the distinct behaviors it should exhibit (happy path, edge cases, error cases)
+3. **Find existing tests** — search for test files covering the changed code. Check:
+   - Test file naming conventions (`.test.ts`, `.spec.ts`, `__tests__/`, etc.)
+   - Import references to the changed modules
+   - Existing test descriptions that mention the changed functionality
+4. **Evaluate coverage gaps**:
+   - Are the new/changed behaviors covered by tests?
+   - Do existing tests still make sense after the changes (stale assertions)?
+   - Are edge cases tested (empty input, boundary values, error conditions)?
+   - Are tests actually asserting the right thing (testing behavior vs testing implementation)?
+5. **Evaluate test quality** for any new/changed tests:
+   - Do test names describe the behavior being tested?
+   - Are assertions specific enough to catch regressions?
+   - Do tests depend on implementation details that could change?
+   - Are there flaky patterns (timing-dependent, order-dependent, global state)?
+
+## Output Format
+
+Report findings as a numbered list. For each finding:
+
+```
+### Finding N: <short title>
+
+**File:** `path/to/file.ts` (function/behavior)
+**Confidence:** <0-100>
+**Severity:** critical | high | medium | low
+**Gap type:** missing-test | weak-assertion | stale-test | flaky-pattern
+
+<Description of what's not tested and why it matters. Include a brief sketch of what a test should verify.>
+```
+
+**Only report findings with confidence ≥ 80.**
+
+If coverage looks solid, say "No findings — test coverage looks adequate." with a summary of what you checked.
+
+## Important
+
+- Don't demand 100% coverage. Focus on **behaviors that matter** — things that could break and affect users.
+- Missing tests for trivial getters/setters are not findings. Missing tests for complex logic or error handling are.
+- If the project has no test infrastructure at all, note that as a single finding rather than listing every untested function.
+- Do NOT use `#N` notation in your output (GitHub auto-links it to issues). Use "finding N" or "item N" instead.
+
+
+## Constraints
+
+- **Never post to GitHub.** Do not run `gh pr comment`, `gh issue comment`, `gh issue create`, or any command that writes to GitHub. Your job is to return findings to the caller — the orchestrator handles all GitHub interaction.

package/dist/bun/cli.d.ts

@@ -0,0 +1,3 @@
+#!/usr/bin/env node
+export {};
+//# sourceMappingURL=cli.d.ts.map

package/dist/bun/cli.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"cli.d.ts","sourceRoot":"","sources":["../../src/bun/cli.ts"],"names":[],"mappings":"","sourcesContent":["#!/usr/bin/env node\nprocess.title = \"dreb\";\nprocess.emitWarning = (() => {}) as typeof process.emitWarning;\n\nawait import(\"./register-bedrock.js\");\nawait import(\"../cli.js\");\n"]}

package/dist/bun/cli.js

@@ -0,0 +1,7 @@
+#!/usr/bin/env node
+process.title = "dreb";
+process.emitWarning = (() => { });
+await import("./register-bedrock.js");
+await import("../cli.js");
+export {};
+//# sourceMappingURL=cli.js.map

package/dist/bun/cli.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"cli.js","sourceRoot":"","sources":["../../src/bun/cli.ts"],"names":[],"mappings":";AACA,OAAO,CAAC,KAAK,GAAG,MAAM,CAAC;AACvB,OAAO,CAAC,WAAW,GAAG,CAAC,GAAG,EAAE,CAAC,EAAC,CAAC,CAA+B,CAAC;AAE/D,MAAM,MAAM,CAAC,uBAAuB,CAAC,CAAC;AACtC,MAAM,MAAM,CAAC,WAAW,CAAC,CAAC","sourcesContent":["#!/usr/bin/env node\nprocess.title = \"dreb\";\nprocess.emitWarning = (() => {}) as typeof process.emitWarning;\n\nawait import(\"./register-bedrock.js\");\nawait import(\"../cli.js\");\n"]}

package/dist/bun/register-bedrock.d.ts

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=register-bedrock.d.ts.map

package/dist/bun/register-bedrock.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"register-bedrock.d.ts","sourceRoot":"","sources":["../../src/bun/register-bedrock.ts"],"names":[],"mappings":"","sourcesContent":["import { setBedrockProviderModule } from \"@dreb/ai\";\nimport { bedrockProviderModule } from \"@dreb/ai/bedrock-provider\";\n\nsetBedrockProviderModule(bedrockProviderModule);\n"]}

package/dist/bun/register-bedrock.js

@@ -0,0 +1,4 @@
+import { setBedrockProviderModule } from "@dreb/ai";
+import { bedrockProviderModule } from "@dreb/ai/bedrock-provider";
+setBedrockProviderModule(bedrockProviderModule);
+//# sourceMappingURL=register-bedrock.js.map

package/dist/bun/register-bedrock.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"register-bedrock.js","sourceRoot":"","sources":["../../src/bun/register-bedrock.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,wBAAwB,EAAE,MAAM,UAAU,CAAC;AACpD,OAAO,EAAE,qBAAqB,EAAE,MAAM,2BAA2B,CAAC;AAElE,wBAAwB,CAAC,qBAAqB,CAAC,CAAC","sourcesContent":["import { setBedrockProviderModule } from \"@dreb/ai\";\nimport { bedrockProviderModule } from \"@dreb/ai/bedrock-provider\";\n\nsetBedrockProviderModule(bedrockProviderModule);\n"]}

package/dist/cli/args.d.ts

@@ -0,0 +1,50 @@
+/**
+ * CLI argument parsing and help display
+ */
+import type { ThinkingLevel } from "@dreb/agent-core";
+import { type ToolName } from "../core/tools/index.js";
+export type Mode = "text" | "json" | "rpc";
+export interface Args {
+    provider?: string;
+    model?: string;
+    apiKey?: string;
+    systemPrompt?: string;
+    appendSystemPrompt?: string;
+    thinking?: ThinkingLevel;
+    continue?: boolean;
+    resume?: boolean;
+    help?: boolean;
+    version?: boolean;
+    mode?: Mode;
+    ui?: string;
+    noSession?: boolean;
+    session?: string;
+    fork?: string;
+    sessionDir?: string;
+    models?: string[];
+    tools?: ToolName[];
+    noTools?: boolean;
+    extensions?: string[];
+    noExtensions?: boolean;
+    print?: boolean;
+    export?: string;
+    noSkills?: boolean;
+    skills?: string[];
+    promptTemplates?: string[];
+    noPromptTemplates?: boolean;
+    themes?: string[];
+    noThemes?: boolean;
+    listModels?: string | true;
+    offline?: boolean;
+    verbose?: boolean;
+    messages: string[];
+    fileArgs: string[];
+    /** Unknown flags (potentially extension flags) - map of flag name to value */
+    unknownFlags: Map<string, boolean | string>;
+}
+export declare function isValidThinkingLevel(level: string): level is ThinkingLevel;
+export declare function parseArgs(args: string[], extensionFlags?: Map<string, {
+    type: "boolean" | "string";
+}>): Args;
+export declare function printHelp(): void;
+//# sourceMappingURL=args.d.ts.map

package/dist/cli/args.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"args.d.ts","sourceRoot":"","sources":["../../src/cli/args.ts"],"names":[],"mappings":"AAAA;;GAEG;AAEH,OAAO,KAAK,EAAE,aAAa,EAAE,MAAM,kBAAkB,CAAC;AAGtD,OAAO,EAAY,KAAK,QAAQ,EAAE,MAAM,wBAAwB,CAAC;AAEjE,MAAM,MAAM,IAAI,GAAG,MAAM,GAAG,MAAM,GAAG,KAAK,CAAC;AAE3C,MAAM,WAAW,IAAI;IACpB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,kBAAkB,CAAC,EAAE,MAAM,CAAC;IAC5B,QAAQ,CAAC,EAAE,aAAa,CAAC;IACzB,QAAQ,CAAC,EAAE,OAAO,CAAC;IACnB,MAAM,CAAC,EAAE,OAAO,CAAC;IACjB,IAAI,CAAC,EAAE,OAAO,CAAC;IACf,OAAO,CAAC,EAAE,OAAO,CAAC;IAClB,IAAI,CAAC,EAAE,IAAI,CAAC;IACZ,EAAE,CAAC,EAAE,MAAM,CAAC;IACZ,SAAS,CAAC,EAAE,OAAO,CAAC;IACpB,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,MAAM,CAAC,EAAE,MAAM,EAAE,CAAC;IAClB,KAAK,CAAC,EAAE,QAAQ,EAAE,CAAC;IACnB,OAAO,CAAC,EAAE,OAAO,CAAC;IAClB,UAAU,CAAC,EAAE,MAAM,EAAE,CAAC;IACtB,YAAY,CAAC,EAAE,OAAO,CAAC;IACvB,KAAK,CAAC,EAAE,OAAO,CAAC;IAChB,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,QAAQ,CAAC,EAAE,OAAO,CAAC;IACnB,MAAM,CAAC,EAAE,MAAM,EAAE,CAAC;IAClB,eAAe,CAAC,EAAE,MAAM,EAAE,CAAC;IAC3B,iBAAiB,CAAC,EAAE,OAAO,CAAC;IAC5B,MAAM,CAAC,EAAE,MAAM,EAAE,CAAC;IAClB,QAAQ,CAAC,EAAE,OAAO,CAAC;IACnB,UAAU,CAAC,EAAE,MAAM,GAAG,IAAI,CAAC;IAC3B,OAAO,CAAC,EAAE,OAAO,CAAC;IAClB,OAAO,CAAC,EAAE,OAAO,CAAC;IAClB,QAAQ,EAAE,MAAM,EAAE,CAAC;IACnB,QAAQ,EAAE,MAAM,EAAE,CAAC;IACnB,8EAA8E;IAC9E,YAAY,EAAE,GAAG,CAAC,MAAM,EAAE,OAAO,GAAG,MAAM,CAAC,CAAC;CAC5C;AAID,wBAAgB,oBAAoB,CAAC,KAAK,EAAE,MAAM,GAAG,KAAK,IAAI,aAAa,CAE1E;AAED,wBAAgB,SAAS,CAAC,IAAI,EAAE,MAAM,EAAE,EAAE,cAAc,CAAC,EAAE,GAAG,CAAC,MAAM,EAAE;IAAE,IAAI,EAAE,SAAS,GAAG,QAAQ,CAAA;CAAE,CAAC,GAAG,IAAI,CA8H5G;AAED,wBAAgB,SAAS,IAAI,IAAI,CA8IhC","sourcesContent":["/**\n * CLI argument parsing and help display\n */\n\nimport type { ThinkingLevel } from \"@dreb/agent-core\";\nimport chalk from \"chalk\";\nimport { APP_NAME, CONFIG_DIR_NAME, ENV_AGENT_DIR } from \"../config.js\";\nimport { allTools, type ToolName } from \"../core/tools/index.js\";\n\nexport type Mode = \"text\" | \"json\" | \"rpc\";\n\nexport interface Args {\n\tprovider?: string;\n\tmodel?: string;\n\tapiKey?: string;\n\tsystemPrompt?: string;\n\tappendSystemPrompt?: string;\n\tthinking?: ThinkingLevel;\n\tcontinue?: boolean;\n\tresume?: boolean;\n\thelp?: boolean;\n\tversion?: boolean;\n\tmode?: Mode;\n\tui?: string;\n\tnoSession?: boolean;\n\tsession?: string;\n\tfork?: string;\n\tsessionDir?: string;\n\tmodels?: string[];\n\ttools?: ToolName[];\n\tnoTools?: boolean;\n\textensions?: string[];\n\tnoExtensions?: boolean;\n\tprint?: boolean;\n\texport?: string;\n\tnoSkills?: boolean;\n\tskills?: string[];\n\tpromptTemplates?: string[];\n\tnoPromptTemplates?: boolean;\n\tthemes?: string[];\n\tnoThemes?: boolean;\n\tlistModels?: string | true;\n\toffline?: boolean;\n\tverbose?: boolean;\n\tmessages: string[];\n\tfileArgs: string[];\n\t/** Unknown flags (potentially extension flags) - map of flag name to value */\n\tunknownFlags: Map<string, boolean | string>;\n}\n\nconst VALID_THINKING_LEVELS = [\"off\", \"minimal\", \"low\", \"medium\", \"high\", \"xhigh\"] as const;\n\nexport function isValidThinkingLevel(level: string): level is ThinkingLevel {\n\treturn VALID_THINKING_LEVELS.includes(level as ThinkingLevel);\n}\n\nexport function parseArgs(args: string[], extensionFlags?: Map<string, { type: \"boolean\" | \"string\" }>): Args {\n\tconst result: Args = {\n\t\tmessages: [],\n\t\tfileArgs: [],\n\t\tunknownFlags: new Map(),\n\t};\n\n\tfor (let i = 0; i < args.length; i++) {\n\t\tconst arg = args[i];\n\n\t\tif (arg === \"--help\" || arg === \"-h\") {\n\t\t\tresult.help = true;\n\t\t} else if (arg === \"--version\" || arg === \"-v\") {\n\t\t\tresult.version = true;\n\t\t} else if (arg === \"--mode\" && i + 1 < args.length) {\n\t\t\tconst mode = args[++i];\n\t\t\tif (mode === \"text\" || mode === \"json\" || mode === \"rpc\") {\n\t\t\t\tresult.mode = mode;\n\t\t\t}\n\t\t} else if (arg === \"--ui\" && i + 1 < args.length) {\n\t\t\tresult.ui = args[++i];\n\t\t} else if (arg === \"--continue\" || arg === \"-c\") {\n\t\t\tresult.continue = true;\n\t\t} else if (arg === \"--resume\" || arg === \"-r\") {\n\t\t\tresult.resume = true;\n\t\t} else if (arg === \"--provider\" && i + 1 < args.length) {\n\t\t\tresult.provider = args[++i];\n\t\t} else if (arg === \"--model\" && i + 1 < args.length) {\n\t\t\tresult.model = args[++i];\n\t\t} else if (arg === \"--api-key\" && i + 1 < args.length) {\n\t\t\tresult.apiKey = args[++i];\n\t\t} else if (arg === \"--system-prompt\" && i + 1 < args.length) {\n\t\t\tresult.systemPrompt = args[++i];\n\t\t} else if (arg === \"--append-system-prompt\" && i + 1 < args.length) {\n\t\t\tresult.appendSystemPrompt = args[++i];\n\t\t} else if (arg === \"--no-session\") {\n\t\t\tresult.noSession = true;\n\t\t} else if (arg === \"--session\" && i + 1 < args.length) {\n\t\t\tresult.session = args[++i];\n\t\t} else if (arg === \"--fork\" && i + 1 < args.length) {\n\t\t\tresult.fork = args[++i];\n\t\t} else if (arg === \"--session-dir\" && i + 1 < args.length) {\n\t\t\tresult.sessionDir = args[++i];\n\t\t} else if (arg === \"--models\" && i + 1 < args.length) {\n\t\t\tresult.models = args[++i].split(\",\").map((s) => s.trim());\n\t\t} else if (arg === \"--no-tools\") {\n\t\t\tresult.noTools = true;\n\t\t} else if (arg === \"--tools\" && i + 1 < args.length) {\n\t\t\tconst toolNames = args[++i].split(\",\").map((s) => s.trim());\n\t\t\tconst validTools: ToolName[] = [];\n\t\t\tfor (const name of toolNames) {\n\t\t\t\tif (name in allTools) {\n\t\t\t\t\tvalidTools.push(name as ToolName);\n\t\t\t\t} else {\n\t\t\t\t\tconsole.error(\n\t\t\t\t\t\tchalk.yellow(`Warning: Unknown tool \"${name}\". Valid tools: ${Object.keys(allTools).join(\", \")}`),\n\t\t\t\t\t);\n\t\t\t\t}\n\t\t\t}\n\t\t\tresult.tools = validTools;\n\t\t} else if (arg === \"--thinking\" && i + 1 < args.length) {\n\t\t\tconst level = args[++i];\n\t\t\tif (isValidThinkingLevel(level)) {\n\t\t\t\tresult.thinking = level;\n\t\t\t} else {\n\t\t\t\tconsole.error(\n\t\t\t\t\tchalk.yellow(\n\t\t\t\t\t\t`Warning: Invalid thinking level \"${level}\". Valid values: ${VALID_THINKING_LEVELS.join(\", \")}`,\n\t\t\t\t\t),\n\t\t\t\t);\n\t\t\t}\n\t\t} else if (arg === \"--print\" || arg === \"-p\") {\n\t\t\tresult.print = true;\n\t\t} else if (arg === \"--export\" && i + 1 < args.length) {\n\t\t\tresult.export = args[++i];\n\t\t} else if ((arg === \"--extension\" || arg === \"-e\") && i + 1 < args.length) {\n\t\t\tresult.extensions = result.extensions ?? [];\n\t\t\tresult.extensions.push(args[++i]);\n\t\t} else if (arg === \"--no-extensions\" || arg === \"-ne\") {\n\t\t\tresult.noExtensions = true;\n\t\t} else if (arg === \"--skill\" && i + 1 < args.length) {\n\t\t\tresult.skills = result.skills ?? [];\n\t\t\tresult.skills.push(args[++i]);\n\t\t} else if (arg === \"--prompt-template\" && i + 1 < args.length) {\n\t\t\tresult.promptTemplates = result.promptTemplates ?? [];\n\t\t\tresult.promptTemplates.push(args[++i]);\n\t\t} else if (arg === \"--theme\" && i + 1 < args.length) {\n\t\t\tresult.themes = result.themes ?? [];\n\t\t\tresult.themes.push(args[++i]);\n\t\t} else if (arg === \"--no-skills\" || arg === \"-ns\") {\n\t\t\tresult.noSkills = true;\n\t\t} else if (arg === \"--no-prompt-templates\" || arg === \"-np\") {\n\t\t\tresult.noPromptTemplates = true;\n\t\t} else if (arg === \"--no-themes\") {\n\t\t\tresult.noThemes = true;\n\t\t} else if (arg === \"--list-models\") {\n\t\t\t// Check if next arg is a search pattern (not a flag or file arg)\n\t\t\tif (i + 1 < args.length && !args[i + 1].startsWith(\"-\") && !args[i + 1].startsWith(\"@\")) {\n\t\t\t\tresult.listModels = args[++i];\n\t\t\t} else {\n\t\t\t\tresult.listModels = true;\n\t\t\t}\n\t\t} else if (arg === \"--verbose\") {\n\t\t\tresult.verbose = true;\n\t\t} else if (arg === \"--offline\") {\n\t\t\tresult.offline = true;\n\t\t} else if (arg.startsWith(\"@\")) {\n\t\t\tresult.fileArgs.push(arg.slice(1)); // Remove @ prefix\n\t\t} else if (arg.startsWith(\"--\") && extensionFlags) {\n\t\t\t// Check if it's an extension-registered flag\n\t\t\tconst flagName = arg.slice(2);\n\t\t\tconst extFlag = extensionFlags.get(flagName);\n\t\t\tif (extFlag) {\n\t\t\t\tif (extFlag.type === \"boolean\") {\n\t\t\t\t\tresult.unknownFlags.set(flagName, true);\n\t\t\t\t} else if (extFlag.type === \"string\" && i + 1 < args.length) {\n\t\t\t\t\tresult.unknownFlags.set(flagName, args[++i]);\n\t\t\t\t}\n\t\t\t}\n\t\t\t// Unknown flags without extensionFlags are silently ignored (first pass)\n\t\t} else if (!arg.startsWith(\"-\")) {\n\t\t\tresult.messages.push(arg);\n\t\t}\n\t}\n\n\treturn result;\n}\n\nexport function printHelp(): void {\n\tconsole.log(`${chalk.bold(APP_NAME)} - AI coding assistant with read, bash, edit, write tools\n\n${chalk.bold(\"Usage:\")}\n  ${APP_NAME} [options] [@files...] [messages...]\n\n${chalk.bold(\"Commands:\")}\n  ${APP_NAME} install <source> [-l]     Install extension source and add to settings\n  ${APP_NAME} remove <source> [-l]      Remove extension source from settings\n  ${APP_NAME} uninstall <source> [-l]   Alias for remove\n  ${APP_NAME} update [source]           Update installed extensions (skips pinned sources)\n  ${APP_NAME} list                      List installed extensions from settings\n  ${APP_NAME} config                    Open TUI to enable/disable package resources\n  ${APP_NAME} <command> --help          Show help for install/remove/uninstall/update/list\n\n${chalk.bold(\"Options:\")}\n  --provider <name>              Provider name (default: google)\n  --model <pattern>              Model pattern or ID (supports \"provider/id\" and optional \":<thinking>\")\n  --api-key <key>                API key (defaults to env vars)\n  --system-prompt <text>         System prompt (default: coding assistant prompt)\n  --append-system-prompt <text>  Append text or file contents to the system prompt\n  --mode <mode>                  Output mode: text (default), json, or rpc\n  --ui <type>                    UI type hint for system prompt (e.g. tui, telegram, rpc)\n  --print, -p                    Non-interactive mode: process prompt and exit\n  --continue, -c                 Continue previous session\n  --resume, -r                   Select a session to resume\n  --session <path>               Use specific session file\n  --fork <path>                  Fork specific session file or partial UUID into a new session\n  --session-dir <dir>            Directory for session storage and lookup\n  --no-session                   Don't save session (ephemeral)\n  --models <patterns>            Comma-separated model patterns for Ctrl+P cycling\n                                 Supports globs (anthropic/*, *sonnet*) and fuzzy matching\n  --no-tools                     Disable all built-in tools\n  --tools <tools>                Comma-separated list of tools to enable (default: all)\n                                 Available: read, bash, edit, write, grep, find, ls, web_search, web_fetch, subagent\n  --thinking <level>             Set thinking level: off, minimal, low, medium, high, xhigh\n  --extension, -e <path>         Load an extension file (can be used multiple times)\n  --no-extensions, -ne           Disable extension discovery (explicit -e paths still work)\n  --skill <path>                 Load a skill file or directory (can be used multiple times)\n  --no-skills, -ns               Disable skills discovery and loading\n  --prompt-template <path>       Load a prompt template file or directory (can be used multiple times)\n  --no-prompt-templates, -np     Disable prompt template discovery and loading\n  --theme <path>                 Load a theme file or directory (can be used multiple times)\n  --no-themes                    Disable theme discovery and loading\n  --export <file>                Export session file to HTML and exit\n  --list-models [search]         List available models (with optional fuzzy search)\n  --verbose                      Force verbose startup (overrides quietStartup setting)\n  --offline                      Disable startup network operations (same as DREB_OFFLINE=1)\n  --help, -h                     Show this help\n  --version, -v                  Show version number\n\nExtensions can register additional flags (e.g., --plan from plan-mode extension).\n\n${chalk.bold(\"Examples:\")}\n  # Interactive mode\n  ${APP_NAME}\n\n  # Interactive mode with initial prompt\n  ${APP_NAME} \"List all .ts files in src/\"\n\n  # Include files in initial message\n  ${APP_NAME} @prompt.md @image.png \"What color is the sky?\"\n\n  # Non-interactive mode (process and exit)\n  ${APP_NAME} -p \"List all .ts files in src/\"\n\n  # Multiple messages (interactive)\n  ${APP_NAME} \"Read package.json\" \"What dependencies do we have?\"\n\n  # Continue previous session\n  ${APP_NAME} --continue \"What did we discuss?\"\n\n  # Use different model\n  ${APP_NAME} --provider openai --model gpt-4o-mini \"Help me refactor this code\"\n\n  # Use model with provider prefix (no --provider needed)\n  ${APP_NAME} --model openai/gpt-4o \"Help me refactor this code\"\n\n  # Use model with thinking level shorthand\n  ${APP_NAME} --model sonnet:high \"Solve this complex problem\"\n\n  # Limit model cycling to specific models\n  ${APP_NAME} --models claude-sonnet,claude-haiku,gpt-4o\n\n  # Limit to a specific provider with glob pattern\n  ${APP_NAME} --models \"github-copilot/*\"\n\n  # Cycle models with fixed thinking levels\n  ${APP_NAME} --models sonnet:high,haiku:low\n\n  # Start with a specific thinking level\n  ${APP_NAME} --thinking high \"Solve this complex problem\"\n\n  # Read-only mode (no file modifications possible)\n  ${APP_NAME} --tools read,grep,find,ls -p \"Review the code in src/\"\n\n  # Export a session file to HTML\n  ${APP_NAME} --export ~/${CONFIG_DIR_NAME}/agent/sessions/--path--/session.jsonl\n  ${APP_NAME} --export session.jsonl output.html\n\n${chalk.bold(\"Environment Variables:\")}\n  ANTHROPIC_API_KEY                - Anthropic Claude API key\n  ANTHROPIC_OAUTH_TOKEN            - Anthropic OAuth token (alternative to API key)\n  OPENAI_API_KEY                   - OpenAI GPT API key\n  AZURE_OPENAI_API_KEY             - Azure OpenAI API key\n  AZURE_OPENAI_BASE_URL            - Azure OpenAI base URL (https://{resource}.openai.azure.com/openai/v1)\n  AZURE_OPENAI_RESOURCE_NAME       - Azure OpenAI resource name (alternative to base URL)\n  AZURE_OPENAI_API_VERSION         - Azure OpenAI API version (default: v1)\n  AZURE_OPENAI_DEPLOYMENT_NAME_MAP - Azure OpenAI model=deployment map (comma-separated)\n  GEMINI_API_KEY                   - Google Gemini API key\n  GROQ_API_KEY                     - Groq API key\n  CEREBRAS_API_KEY                 - Cerebras API key\n  XAI_API_KEY                      - xAI Grok API key\n  OPENROUTER_API_KEY               - OpenRouter API key\n  AI_GATEWAY_API_KEY               - Vercel AI Gateway API key\n  ZAI_API_KEY                      - ZAI API key\n  MISTRAL_API_KEY                  - Mistral API key\n  MINIMAX_API_KEY                  - MiniMax API key\n  OPENCODE_API_KEY                 - OpenCode Zen/OpenCode Go API key\n  KIMI_API_KEY                     - Kimi For Coding API key\n  AWS_PROFILE                      - AWS profile for Amazon Bedrock\n  AWS_ACCESS_KEY_ID                - AWS access key for Amazon Bedrock\n  AWS_SECRET_ACCESS_KEY            - AWS secret key for Amazon Bedrock\n  AWS_BEARER_TOKEN_BEDROCK         - Bedrock API key (bearer token)\n  AWS_REGION                       - AWS region for Amazon Bedrock (e.g., us-east-1)\n  ${ENV_AGENT_DIR.padEnd(32)} - Session storage directory (default: ~/${CONFIG_DIR_NAME}/agent)\n  DREB_PACKAGE_DIR                   - Override package directory (for Nix/Guix store paths)\n  DREB_OFFLINE                       - Disable startup network operations when set to 1/true/yes\n  DREB_AI_ANTIGRAVITY_VERSION        - Override Antigravity User-Agent version (e.g., 1.23.0)\n\n${chalk.bold(\"Available Tools (default: all):\")}\n  read       - Read file contents\n  bash       - Execute bash commands\n  edit       - Edit files with find/replace\n  write      - Write files (creates/overwrites)\n  grep       - Search file contents\n  find       - Find files by glob pattern\n  ls         - List directory contents\n  web_search - Search the web\n  web_fetch  - Fetch URL content\n  subagent   - Delegate tasks to independent subagents\n`);\n}\n"]}