diff --git a/public/images/docs/falcon-ai/guides/connect-mcp-server-discovered.png b/public/images/docs/falcon-ai/guides/connect-mcp-server-discovered.png new file mode 100644 index 00000000..ff001d35 Binary files /dev/null and b/public/images/docs/falcon-ai/guides/connect-mcp-server-discovered.png differ diff --git a/public/images/docs/simulation/guides/fix-my-agent.png b/public/images/docs/simulation/guides/fix-my-agent.png new file mode 100644 index 00000000..11680a94 Binary files /dev/null and b/public/images/docs/simulation/guides/fix-my-agent.png differ diff --git a/src/pages/docs/falcon-ai/guides/connect-mcp-server.mdx b/src/pages/docs/falcon-ai/guides/connect-mcp-server.mdx index e79ede49..92c8a8e0 100644 --- a/src/pages/docs/falcon-ai/guides/connect-mcp-server.mdx +++ b/src/pages/docs/falcon-ai/guides/connect-mcp-server.mdx @@ -47,10 +47,13 @@ If you set **Authentication** to **OAuth 2.0**, saving the connector doesn't sig ## Test the connection and discover its tools -With the connector saved and, for OAuth, signed in, select the new connector in the list on the left to open its detail pane. This is where **Test Connection**, **Discover Tools**, **Re-authenticate**, **Edit**, and **Delete** all live. Click **Test Connection** to confirm Falcon AI can reach it. While it runs, the button reads **Testing...**; it resolves to **Connection test succeeded.** or **Connection test failed.** +With the connector saved and, for OAuth, signed in, select the new connector in the list on the left to open its detail pane. This is where **Edit**, **Discover Tools**, **Test Connection**, and **Delete** all live, plus **Re-authenticate** on a connector that signs in through OAuth. Click **Test Connection** to confirm Falcon AI can reach it. While it runs, the button reads **Testing...**; it resolves to **Connection test succeeded.** or **Connection test failed.** Once the connection is good, click **Discover Tools** (it reads **Discovering...** while it runs) to have the server report what it offers. A successful pass shows something like **Discovered 4 tools.** A failed one shows **Tool discovery failed.** +A connected connector's detail pane with the Edit, Discover Tools, Test Connection, and Delete buttons above a Discovered Tools list holding one tool, its description, and an enabled toggle +*Discovery stores what it found against the connector, so the list is still there when you come back without re-running it* + Either failure points back at the form rather than at Falcon AI, so click **Edit** and re-check three things: the **Server URL** the server actually listens on, the **Transport** it expects (an older server given Streamable HTTP fails here rather than at save), and, for API Key, that **Header Name** matches what the server looks for and not just the credential in **Header Value**. Save and run **Test Connection** again. ## Narrow down the tools diff --git a/src/pages/docs/simulation/guides/fix-my-agent.mdx b/src/pages/docs/simulation/guides/fix-my-agent.mdx index 8e18ce96..8c378929 100644 --- a/src/pages/docs/simulation/guides/fix-my-agent.mdx +++ b/src/pages/docs/simulation/guides/fix-my-agent.mdx @@ -7,7 +7,7 @@ description: "Turn a finished run into a ranked list of issues and fixes, then h ## Open it from a finished run -On a run's [results page](/docs/simulation/guides/explore-results), **Fix My Agent** sits next to the tabs rather than inside one of them. Click it to open a side panel that stays open alongside whichever tab you're on, with a chevron to collapse it out of the way when you don't need it. +On a run's [results page](/docs/simulation/guides/explore-results), **Fix My Agent** sits next to the tabs rather than inside one of them. Click it to open a side panel that stays open alongside whichever tab you're on, with a chevron at the panel's left edge to collapse it out of the way when you don't need it. The button only turns on once the run has enough to analyse: the run has to be completed, and it needs at least 15 connected calls behind it. A run still in progress, or one with fewer calls than that, leaves the button disabled, with a tooltip telling you which of the two is missing. @@ -15,16 +15,23 @@ The button only turns on once the run has enough to analyse: the run has to be c The first time you open the panel on a run, it's empty: "There are no suggestions yet, click the refresh button to get suggestions." Click refresh to run the analysis over the run's calls. If it genuinely finds nothing worth flagging, it says so instead of manufacturing an issue to fill the space. +An analysis is kept against the execution it ran on, stamped **Last updated at** beside the suggestion count, so reopening the panel later shows you that stored result rather than starting again. That's also why the empty state is a first-time-only thing: once a run has been analysed, it's the stored analysis you come back to, and refresh is what replaces it. + ## What a prioritised issue looks like Each entry in the list is one issue, not one call. A run where a dozen calls fail the same way for the same reason surfaces as a single entry, not a dozen. Every entry carries: - A short **heading** naming the issue - A **priority**, high, medium, or low, so you know which to read first -- A written **recommendation** of what to change to address it -- The **calls it's drawn from**. Click the entry and the calls grid on the page narrows to just those, so you can read the transcripts behind the pattern before you act on it +- A written **recommendation** of what to change to address it, truncated behind a **see more** link +- The **calls it's drawn from**, as **Calls Affected (n)**. Click the entry and the calls grid on the page narrows to just those, its header switching from **All Chats** to **(n) Chats selected**, so you can read the transcripts behind the pattern before you act on it + +The Fix My Agent panel open beside the Chat Details tab, showing the run summary, the Agent Level and Branch Level toggle, and two branch-level issues each with a priority chip, a recommendation, a branch category, and Calls Affected +*The panel opens over the right of the page and stays there while you move between tabs, which is how you read an issue and its calls side by side* + +The group is headed **Suggestions (n)**, the number the optimizer can act on, with a **Summary** of what the run showed overall underneath it. Below that, **Actionable Suggestions** splits the same set across two tabs: **Agent Level**, for issues with your agent's prompt as a whole, and **Branch Level**, for issues tied to one path through the conversation, which an entry names on a **Branch Category** line. The per-tab count beside them tells you how many of the total you're currently looking at, so the two tabs always sum to the heading. -The list carries two counts: how many issues were found in total, and how many of those are written as a fix to your agent's prompt, the ones you can act on directly. You can switch between issues at the agent level and issues at the domain level. A separate group covers issues that aren't about the prompt at all, things about how the agent is set up rather than what it says. Those are worth reading, but they're informational: an optimization run can't act on them the way it can on a prompt-based suggestion. +**Infra based suggestions** is a separate group, for what the run showed about the agent's runtime rather than its prompt: response latency, timeouts, conversations that loop. Entries look the same as the prompt-based ones, priority and recommendation and the calls behind them, but as the group says, they aren't supported by the optimizer and you make those updates manually. The group also opens with a **Human Comparison Summary**, which reads the run's latency, turn count, and CSAT against what a human agent typically achieves. ## Hand off to an optimization