The two new retrieval bars get their Settings controls #154

Merged

bvandeusen merged 1 commits from dev into main 2026-09-11 20:58:50 -04:00

Conversation 0 Commits 1 Files Changed 2

+162 -8

bvandeusen commented 2026-09-11 20:58:44 -04:00

Owner

Copy Link

Copy Source

One commit. #3852 and #3853 each added a threshold and neither added its input, so two of the eleven retrieval settings were reachable only through the database.

What ships

• kb_toolrule_threshold — the command arm's bar. Its hint says why it sits below the write-path one rather than leaving that looking like a mistake: a shell command is short, so it scores lower for the same relevance, and at a shared bar this arm spoke on 2% of calls against the write path's 37%.
• kb_promptrule_threshold — the prompt boundary, a third query shape again, and the only moment that reaches a rule about how to answer rather than how to act.

Both follow the five-site pattern the existing nine controls use (ref → clamp on save → write-back → payload key → load). Separate keys on purpose: the finding of #3853 is that one number cannot serve arms whose queries differ in shape.

Also corrects copy that went stale hours earlier. The standing-rule hint still said the arm surfaces "only rules marked conditional, since always-on ones are already loaded" — describing a tier milestone 394 removed, on the one surface whose job is telling the operator what the bar does.

The guard

A form's initial value is a claim about the server's default, and nothing connected the two. Retune the Python constant and the input keeps rendering the old number, which an operator who has never touched it reads as the bar in force. Nothing errors; the form simply misinforms them about the one fact it exists to convey.

tests/test_settings_defaults_agree.py pins that relationship across all five thresholds — never the values, so retuning stays free as long as both move. Falsified against a drifted form value before commit.

A correction worth recording

The issue this closes claimed ten settings had no UI. That was wrong, and I wrote it: the evidence was a grep across web/src, a directory this repo does not have. Nine of eleven had controls all along.

An empty result from an impossible query is indistinguishable from an empty result from a real one — #3720's defect, committed while filing issues about the product doing the same thing. The issue body carries the correction rather than being quietly rewritten.

The second half of that issue — a REST route for retrieval_telemetry plus near-miss distributions beside each bar — was declined by the operator and is not built.

All six CI jobs green on 2e8d846, first run.

🤖 Generated with Claude Code

https://claude.ai/code/session_011cPyzNnegXHr5iRMzzy5KJ

One commit. #3852 and #3853 each added a threshold and neither added its input, so two of the eleven retrieval settings were reachable only through the database. ## What ships - **`kb_toolrule_threshold`** — the command arm's bar. Its hint says why it sits *below* the write-path one rather than leaving that looking like a mistake: a shell command is short, so it scores lower for the same relevance, and at a shared bar this arm spoke on 2% of calls against the write path's 37%. - **`kb_promptrule_threshold`** — the prompt boundary, a third query shape again, and the only moment that reaches a rule about how to *answer* rather than how to act. Both follow the five-site pattern the existing nine controls use (ref → clamp on save → write-back → payload key → load). Separate keys on purpose: the finding of #3853 is that one number cannot serve arms whose queries differ in shape. **Also corrects copy that went stale hours earlier.** The standing-rule hint still said the arm surfaces "only rules marked *conditional*, since always-on ones are already loaded" — describing a tier milestone 394 removed, on the one surface whose job is telling the operator what the bar does. ## The guard A form's initial value is a **claim about the server's default**, and nothing connected the two. Retune the Python constant and the input keeps rendering the old number, which an operator who has never touched it reads as the bar in force. Nothing errors; the form simply misinforms them about the one fact it exists to convey. `tests/test_settings_defaults_agree.py` pins that relationship across all five thresholds — never the values, so retuning stays free as long as both move. Falsified against a drifted form value before commit. ## A correction worth recording The issue this closes claimed **ten** settings had no UI. That was wrong, and I wrote it: the evidence was a grep across `web/src`, a directory this repo does not have. Nine of eleven had controls all along. An empty result from an impossible query is indistinguishable from an empty result from a real one — #3720's defect, committed while filing issues about the product doing the same thing. The issue body carries the correction rather than being quietly rewritten. The second half of that issue — a REST route for `retrieval_telemetry` plus near-miss distributions beside each bar — was declined by the operator and is **not built**. All six CI jobs green on `2e8d846`, first run. 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_011cPyzNnegXHr5iRMzzy5KJ

bvandeusen added 1 commit 2026-09-11 20:58:44 -04:00

feat(settings): the two new retrieval bars get their controls (#3927) ... 2e8d8461cc

#3852 and #3853 each added a threshold and neither added its input, so two
of the eleven retrieval settings were reachable only through the database.
The other nine have had UI all along.

That matters because the issue said otherwise. #3927 claimed none of the ten
appeared in the frontend, on the strength of a grep across `web/src` — a
directory this repo does not have. An empty result from a path that cannot
match was read as "no UI anywhere", and a rule-25 argument was written on top
of it. The issue is corrected rather than quietly rewritten: it is #3720's
defect, absence read as non-existence, committed while filing issues about
the product doing the same thing. A grep that returns nothing and a grep that
cannot match produce the same output, and only a positive control tells them
apart.

So this is small, which is the honest size:

- `kb_toolrule_threshold` — the command arm's bar. The hint says why it sits
  BELOW the write-path one rather than leaving that looking like a mistake: a
  shell command is short, so it scores lower for the same relevance, and at a
  shared bar this arm spoke on 2% of calls against the write path's 37%.
- `kb_promptrule_threshold` — the prompt boundary, a third query shape again,
  and the only moment that reaches a rule about how to ANSWER.

Both follow the five-site pattern the existing controls use: ref, clamp on
save, write-back, payload key, load. Separate keys, because the finding of
#3853 is that one number cannot serve arms whose queries differ in shape.

Also corrects copy that went stale this morning. The standing-rule hint still
said the arm surfaces "only rules marked conditional, since always-on ones
are already loaded" — describing a tier milestone 394 removed, on the surface
whose whole job is telling the operator what the bar does.

The guard is the part worth keeping. A form's initial value is a CLAIM about
the server's default, and nothing connected the two: retune the Python
constant and the input keeps rendering the old number, which the operator
reads as the bar in force. It pins the relationship across all five
thresholds, never the values, so retuning stays free as long as both move —
and it is falsified against a drifted form value.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_011cPyzNnegXHr5iRMzzy5KJ

bvandeusen merged commit b7e648226b into main 2026-09-11 20:58:50 -04:00

Sign in to join this conversation.

Reviewers

No Reviewers

Labels

No items

No labels

Milestone

No items

No Milestone

Projects

Clear projects

No projects

Assignees

Clear assignees

renovate-bot bvandeusen

No Assignees

1 Participants

Notifications

Subscribe

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: bvandeusen/FabledScribe#154