docs: 0.5.0 README, and remove the prompt library from the front end - #13
Merged
Conversation
## Removing the prompt library The prompt library had already stopped doing anything. `resolve_app_prompt_body` dropped the legacy `prompt_id` fallback on 2026-07-31, so a digital employee's system prompt comes from `agent_md` and nothing else. Checked before deleting: nine apps still carry a `prompt_id`, and all nine already have an `agent_md`. The chat runtime does not read the prompts table at all. So the menu entry was pointing at a feature that no longer had an effect, with a "being retired" notice attached. Gone now, along with the pages, the API proxy routes, the `prompts` i18n namespace, and the `prompt_id` field that was being threaded through app types and payloads. `/monitoring/prompts` is gone too — it was only ever a title mapping in `DynamicTitle`; the page itself never existed, so that entry pointed at a 404. `AgentMdEditor` no longer takes `promptId`. It only drove wording — "upgrade to Agent.md" versus "create Agent.md" — and with no prompt library there is nothing to upgrade from, so it always says create. Deliberately kept, since they only share the word "prompt": the workflow node's `systemPrompt` field, the dataset `promptType` setting, the crawler's prompt, and `/api/v1/prompt-variables`, which the skill editor uses. The backend `prompts` table and its endpoints are untouched — the public chat API still accepts an explicit `prompt_id`, and removing that is a separate decision made on the service side. ## README Switched to the new dashboard screenshot, which shows Digital Employees and Skills. The caption says plainly that the greyed-out Prompt entry and the version badge in the image predate this release — the image was taken before these changes, and a screenshot showing a menu item that no longer exists is exactly the kind of small wrongness that confuses a new reader. Rewrote the prompts section from "being retired" to "removed", and added the built-in skill-creator to the Skills section: it is the headline of 0.5.0 and the README did not mention it. The paragraph leads with what it does and ends with what it cannot do — approval stays human, and a draft's scripts are not executable until someone approves them. ## Checks 246 passed. `tsc` reports 134 errors against a 135 baseline — one fewer, since the deleted code carried one. `check:i18n` still fails, as it does on `main`: 3676 reported "missing keys" there versus 3592 here, almost all of them URL strings the scanner mistakes for keys. Pre-existing and unrelated. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Removing the prompt library
The prompt library had already stopped doing anything.
resolve_app_prompt_bodydropped the legacy
prompt_idfallback on 2026-07-31, so a digital employee's systemprompt comes from
agent_mdand nothing else.Checked before deleting rather than assumed:
prompt_idagent_mdSo the menu entry pointed at a feature that no longer had an effect, with a "being
retired" notice attached to it. Removed: the pages, the API proxy routes, the
promptsi18n namespace, the sidebar entry, and theprompt_idfield that was beingthreaded through app types and payloads.
/monitoring/promptswent too. It was only ever a title mapping inDynamicTitle—the page itself never existed, so the entry pointed at a 404.
AgentMdEditorno longer takespromptId. It only drove wording — "upgrade toAgent.md" versus "create Agent.md" — and with no prompt library there is nothing to
upgrade from, so it always says create.
Deliberately kept, since they only share the word prompt: the workflow node's
systemPromptfield, the datasetpromptTypesetting, the crawler's prompt, and/api/v1/prompt-variables, which the skill editor uses.The backend
promptstable and its endpoints are untouched — the public chat APIstill accepts an explicit
prompt_id, and retiring that is a separate decision madeon the service side.
README
Switched to the new dashboard screenshot, which shows Digital Employees and Skills.
The caption states plainly that the greyed-out Prompt entry and the version badge
visible in the image predate this release. The screenshot was taken before these
changes, and a README hero image showing a menu item that no longer exists is exactly
the kind of small wrongness that confuses a new reader. Worth retaking once this
merges, after which the caveat can go.
Rewrote the prompts section from "being retired" to "removed", and added the built-in
skill-creator to the Skills section — it is the headline of 0.5.0 and the README did
not mention it. That paragraph leads with what it does and closes with what it cannot
do: approval stays a human action, and a draft's scripts are not in the executable set
until someone approves them.
Checks
246 passed.
tscreports 134 errors against a 135 baseline — one fewer, since thedeleted code carried one.
check:i18nstill fails, exactly as it does onmain: 3676 reported "missing keys"there versus 3592 here, almost all of them URL strings the scanner mistakes for keys.
Pre-existing and unrelated to this change.
Net −1405 lines.
🤖 Generated with Claude Code