Skip to content

Commit c9104ce

Browse files
authored
Merge pull request #166 from nilenso/docs/fix-pi-ai-package-path
docs: fix pi-ai package path in upstream links
2 parents 16b50e3 + f5916ca commit c9104ce

3 files changed

Lines changed: 6 additions & 6 deletions

File tree

README.md

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -17,7 +17,7 @@ Megasthenes allows you to programmatically ask questions to a GitHub/GitLab repo
1717

1818
- 🔗 **Ask questions about any GitHub/GitLab repository**: Point it at any public or private repository URL and start asking questions in plain language.
1919
- 📌 **Query any point in history**: Pin your question to a specific branch, tag, or commit
20-
- 🤖 **Configurable**: Choose any model and provider supported by [`pi-ai`](https://github.com/badlogic/pi-mono/blob/main/packages/pi-ai/src/models.generated.ts) (OpenRouter, Anthropic, Google, and more). Customize the system prompt, tool iteration limits, and context compaction settings.
20+
- 🤖 **Configurable**: Choose any model and provider supported by [`pi-ai`](https://github.com/badlogic/pi-mono/blob/main/packages/ai/src/models.generated.ts) (OpenRouter, Anthropic, Google, and more). Customize the system prompt, tool iteration limits, and context compaction settings.
2121
- 🔒 **Sandboxed execution**: Run tool execution (file reads, code search) in an isolated container for exploring untrusted repositories safely
2222
- 📊 **Rich answer metadata**: Every response includes token usage, timing, and a complete record of all tool calls the model made.
2323
- 📡 **OpenTelemetry observability**: Built-in tracing with GenAI semantic conventions — send spans to Langfuse, Jaeger, or any OTel-compatible backend. Zero overhead when no SDK is installed.

docs/src/content/docs/guides/api-keys-and-providers.md

Lines changed: 4 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -43,8 +43,8 @@ await session.ask("Summarize the architecture.", {
4343

4444
`maxIterations` and `thinking` can be overridden on a single `ask()` the same way — see [`AskOptions`][askopts].
4545

46-
[pi-ai]: https://github.com/badlogic/pi-mono/tree/main/packages/pi-ai
47-
[models]: https://github.com/badlogic/pi-mono/blob/main/packages/pi-ai/src/models.generated.ts
48-
[pi-env]: https://github.com/badlogic/pi-mono/tree/main/packages/pi-ai#environment-variables-nodejs-only
49-
[pi-oauth]: https://github.com/badlogic/pi-mono/tree/main/packages/pi-ai#oauth-providers
46+
[pi-ai]: https://github.com/badlogic/pi-mono/tree/main/packages/ai
47+
[models]: https://github.com/badlogic/pi-mono/blob/main/packages/ai/src/models.generated.ts
48+
[pi-env]: https://github.com/badlogic/pi-mono/tree/main/packages/ai#environment-variables-nodejs-only
49+
[pi-oauth]: https://github.com/badlogic/pi-mono/tree/main/packages/ai#oauth-providers
5050
[askopts]: /megasthenes/api/interfaces/askoptions/

docs/src/content/docs/start-here/getting-started.md

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -11,7 +11,7 @@ megasthenes is a TypeScript library that lets you programmatically ask questions
1111

1212
- **Ask questions about any repository** — Point it at any public or private repository URL and start asking questions in plain language.
1313
- **Query any point in history** — Pin your question to a specific branch, tag, or commit.
14-
- **Configurable** — Choose any model and provider supported by [pi-ai](https://github.com/badlogic/pi-mono/blob/main/packages/pi-ai/src/models.generated.ts) (OpenRouter, Anthropic, Google, and more). Customize the system prompt, tool iteration limits, and context compaction settings.
14+
- **Configurable** — Choose any model and provider supported by [pi-ai](https://github.com/badlogic/pi-mono/blob/main/packages/ai/src/models.generated.ts) (OpenRouter, Anthropic, Google, and more). Customize the system prompt, tool iteration limits, and context compaction settings.
1515
- **Sandboxed execution** — Run tool execution in an isolated container for exploring untrusted repositories safely.
1616
- **Rich answer metadata** — Every response includes token usage, timing, and a complete record of all tool calls the model made.
1717
- **OpenTelemetry observability** — All LLM calls and tool invocations are traced with GenAI semantic conventions.

0 commit comments

Comments
 (0)