Skip to content

Commit 14a056a

Browse files
committed
docs: fix wrong spacings and stale references in CLI agents toolbox page
1 parent 3dd4c87 commit 14a056a

4 files changed

Lines changed: 4 additions & 8 deletions

File tree

‎website/docs/developer-tools/cli-coding-agents.md‎

Lines changed: 1 addition & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,5 @@
11
---
22
title: CLI Coding Agents
3-
sidebar_position: 1
43
---
54

65
import ToolMark from '@site/src/components/VisualElements/ToolMark';
@@ -15,7 +14,7 @@ This is an opinionated shortlist for serious engineering work, ordered by recomm
1514

1615
Every harness spends a finite budget — tokens, context, attention — on operating itself: system prompts, tool definitions, MCP servers, permission flows, and feature surface. **The more a harness invests in polish, features, and DX, the less of that budget is left for the work itself.** The leaner the harness, the more of the model's capacity goes to your task — but the less it steers the model.
1716

18-
That steering is the other side of the axis. A harness that controls the model heavily shields you from its raw behavior; a lean one exposes it. That is not a defect: LLMs do not reason linearly, and a model can appear to "think wrong" — wander, hedge, revisit — while still converging on the right solution, because its reasoning is probability imitating logic rather than logic itself ([how LLMs work](/how-llms-work#probability-is-not-logic)). The more you let the harness go, the more of that behavior you see.
17+
That steering is the other side of the axis. A harness that controls the model heavily shields you from its raw behavior; a lean one exposes it. That is not a defect: LLMs do not reason linearly, and a model can appear to "think wrong" — wander, hedge, revisit — while still converging on the right solution, because its reasoning is probability imitating logic rather than logic itself ([why LLMs fail](/llm-reliability-limits#probability-is-not-logic)). The more you let the harness go, the more of that behavior you see.
1918

2019
## <span className="rank-numeral">01</span> <ToolMark src="/img/cli-agent-logos/pi.svg" /> pi {#pi}
2120

‎website/docs/developer-tools/cli-tools.md‎

Lines changed: 1 addition & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,5 @@
11
---
22
title: Modern CLI Tools
3-
sidebar_position: 3
43
---
54

65
# Modern CLI Tools for Multi-Agent Workflows
@@ -412,6 +411,6 @@ agent-browser's ref-based approach (`@e1`, `@e2`) produces deterministic element
412411

413412
**Related Course Content:**
414413

415-
- [Chapter 7: Reliability Levers](/reliability-levers) - Multi-worktree workflows leveraging these CLI tools
414+
- [Reliability Levers](/reliability-levers) - Multi-worktree workflows leveraging these CLI tools
416415
- [Developer Tools: Terminals](/developer-tools/terminals) - Terminal recommendations for running these CLI tools efficiently
417416
- [Developer Tools: MCP Servers](/developer-tools/mcp-servers) - Extend CLI agents with code research and web grounding

‎website/docs/developer-tools/mcp-servers.md‎

Lines changed: 2 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,5 @@
11
---
22
title: MCP Servers
3-
sidebar_position: 4
43
---
54

65
# MCP Servers
@@ -37,7 +36,7 @@ uv tool install chunkhound
3736

3837
Requires Python 3.10+ and the uv package manager. See [ChunkHound on GitHub](https://github.com/chunkhound/chunkhound) for API key configuration and setup details.
3938

40-
ChunkHound also handles web research — searching documentation, CVEs, changelogs, and current information beyond training data. Multi-source grounding is described in [Chapter 4](/high-level-methodology#phase-1-grounding).
39+
ChunkHound also handles web research — searching documentation, CVEs, changelogs, and current information beyond training data. Multi-source grounding is described in [Grounding](/workflow-grounding#phase-1-grounding).
4140

4241
## Browser Automation
4342

@@ -60,4 +59,4 @@ Previous recommendations included Playwright MCP and Chrome DevTools MCP. These
6059

6160
**Related Course Content:**
6261

63-
- [Chapter 6: Reliability Levers](/reliability-levers) - Multi-agent workflows that leverage MCP capabilities
62+
- [Reliability Levers](/reliability-levers) - Multi-agent workflows that leverage MCP capabilities

‎website/docs/developer-tools/terminals.md‎

Lines changed: 0 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,5 @@
11
---
22
title: Terminals
3-
sidebar_position: 2
43
---
54

65
import ToolMark from '@site/src/components/VisualElements/ToolMark';

0 commit comments

Comments
 (0)