@codyswann/lisa 4.63.0 → 4.64.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/all/copy-contents/.gitattributes +1 -0
- package/all/copy-contents/gitignore +1 -0
- package/dist/cli/effectiveness-cmd.d.ts +8 -0
- package/dist/cli/effectiveness-cmd.d.ts.map +1 -0
- package/dist/cli/effectiveness-cmd.js +40 -0
- package/dist/cli/effectiveness-cmd.js.map +1 -0
- package/dist/cli/index.d.ts.map +1 -1
- package/dist/cli/index.js +2 -0
- package/dist/cli/index.js.map +1 -1
- package/dist/cli/update-check-hook.js +1 -0
- package/dist/cli/update-check-hook.js.map +1 -1
- package/dist/core/effectiveness-store.d.ts +104 -0
- package/dist/core/effectiveness-store.d.ts.map +1 -0
- package/dist/core/effectiveness-store.js +237 -0
- package/dist/core/effectiveness-store.js.map +1 -0
- package/dist/core/instruction-files-migration.d.ts +2 -1
- package/dist/core/instruction-files-migration.d.ts.map +1 -1
- package/dist/core/instruction-files-migration.js +5 -1
- package/dist/core/instruction-files-migration.js.map +1 -1
- package/dist/core/learnings-merge-driver.d.ts.map +1 -1
- package/dist/core/learnings-merge-driver.js +1 -0
- package/dist/core/learnings-merge-driver.js.map +1 -1
- package/dist/core/nightly-e2e-guard-behavior-certificate.js +2 -2
- package/dist/core/upstream-evidence-manifest.d.ts.map +1 -1
- package/dist/core/upstream-evidence-manifest.js +21 -8
- package/dist/core/upstream-evidence-manifest.js.map +1 -1
- package/dist/utils/effectiveness.d.ts +72 -0
- package/dist/utils/effectiveness.d.ts.map +1 -0
- package/dist/utils/effectiveness.js +183 -0
- package/dist/utils/effectiveness.js.map +1 -0
- package/dist/utils/usage-accounting.d.ts +2 -0
- package/dist/utils/usage-accounting.d.ts.map +1 -1
- package/dist/utils/usage-accounting.js +24 -2
- package/dist/utils/usage-accounting.js.map +1 -1
- package/package.json +4 -4
- package/plugins/lisa/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa/.codex-plugin/plugin.json +1 -1
- package/plugins/lisa/.codex-plugin/skills/lisa-learnings-audit/SKILL.md +10 -1
- package/plugins/lisa/.codex-plugin/skills/lisa-queue-status/SKILL.md +10 -0
- package/plugins/lisa/.codex-plugin/skills/lisa-usage-accounting/SKILL.md +9 -0
- package/plugins/lisa/.codex-plugin/skills/lisa-usage-accounting/references/effectiveness.md +63 -0
- package/plugins/lisa/rules/eager/tool-access-gate.md +13 -11
- package/plugins/lisa/rules/reference/tool-access-gate.md +31 -38
- package/plugins/lisa/rules/reference/usage-accounting.md +5 -1
- package/plugins/lisa/skills/lisa-learnings-audit/SKILL.md +10 -1
- package/plugins/lisa/skills/lisa-queue-status/SKILL.md +10 -0
- package/plugins/lisa/skills/lisa-usage-accounting/SKILL.md +9 -0
- package/plugins/lisa/skills/lisa-usage-accounting/references/effectiveness.md +63 -0
- package/plugins/lisa-agy/plugin.json +1 -1
- package/plugins/lisa-agy/skills/lisa-learnings-audit/SKILL.md +10 -1
- package/plugins/lisa-agy/skills/lisa-queue-status/SKILL.md +10 -0
- package/plugins/lisa-agy/skills/lisa-usage-accounting/SKILL.md +9 -0
- package/plugins/lisa-agy/skills/lisa-usage-accounting/references/effectiveness.md +63 -0
- package/plugins/lisa-cdk/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-cdk/.codex-plugin/plugin.json +1 -1
- package/plugins/lisa-cdk-agy/plugin.json +1 -1
- package/plugins/lisa-cdk-copilot/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-cdk-cursor/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-copilot/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-copilot/rules/eager/tool-access-gate.md +13 -11
- package/plugins/lisa-copilot/rules/reference/tool-access-gate.md +31 -38
- package/plugins/lisa-copilot/rules/reference/usage-accounting.md +5 -1
- package/plugins/lisa-copilot/skills/lisa-learnings-audit/SKILL.md +10 -1
- package/plugins/lisa-copilot/skills/lisa-queue-status/SKILL.md +10 -0
- package/plugins/lisa-copilot/skills/lisa-usage-accounting/SKILL.md +9 -0
- package/plugins/lisa-copilot/skills/lisa-usage-accounting/references/effectiveness.md +63 -0
- package/plugins/lisa-cursor/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-cursor/rules/tool-access-gate-reference.mdc +31 -38
- package/plugins/lisa-cursor/rules/tool-access-gate.mdc +13 -11
- package/plugins/lisa-cursor/rules/usage-accounting-reference.mdc +5 -1
- package/plugins/lisa-cursor/skills/lisa-learnings-audit/SKILL.md +10 -1
- package/plugins/lisa-cursor/skills/lisa-queue-status/SKILL.md +10 -0
- package/plugins/lisa-cursor/skills/lisa-usage-accounting/SKILL.md +9 -0
- package/plugins/lisa-cursor/skills/lisa-usage-accounting/references/effectiveness.md +63 -0
- package/plugins/lisa-expo/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-expo/.codex-plugin/plugin.json +1 -1
- package/plugins/lisa-expo-agy/plugin.json +1 -1
- package/plugins/lisa-expo-copilot/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-expo-cursor/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-harper-fabric/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-harper-fabric/.codex-plugin/plugin.json +1 -1
- package/plugins/lisa-harper-fabric-agy/plugin.json +1 -1
- package/plugins/lisa-harper-fabric-copilot/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-harper-fabric-cursor/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-nestjs/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-nestjs/.codex-plugin/plugin.json +1 -1
- package/plugins/lisa-nestjs-agy/plugin.json +1 -1
- package/plugins/lisa-nestjs-copilot/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-nestjs-cursor/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-openclaw/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-openclaw/.codex-plugin/plugin.json +1 -1
- package/plugins/lisa-openclaw-agy/plugin.json +1 -1
- package/plugins/lisa-openclaw-copilot/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-openclaw-cursor/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-phaser/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-phaser/.codex-plugin/plugin.json +1 -1
- package/plugins/lisa-phaser-agy/plugin.json +1 -1
- package/plugins/lisa-phaser-copilot/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-phaser-cursor/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-rails/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-rails/.codex-plugin/plugin.json +1 -1
- package/plugins/lisa-rails-agy/plugin.json +1 -1
- package/plugins/lisa-rails-copilot/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-rails-cursor/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-typescript/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-typescript/.codex-plugin/plugin.json +1 -1
- package/plugins/lisa-typescript-agy/plugin.json +1 -1
- package/plugins/lisa-typescript-copilot/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-typescript-cursor/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-wiki/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-wiki/.codex-plugin/plugin.json +1 -1
- package/plugins/lisa-wiki-agy/plugin.json +1 -1
- package/plugins/lisa-wiki-copilot/.claude-plugin/plugin.json +1 -1
- package/plugins/lisa-wiki-cursor/.claude-plugin/plugin.json +1 -1
- package/plugins/src/base/rules/eager/tool-access-gate.md +13 -11
- package/plugins/src/base/rules/reference/tool-access-gate.md +31 -38
- package/plugins/src/base/rules/reference/usage-accounting.md +5 -1
- package/plugins/src/base/skills/lisa-learnings-audit/SKILL.md +10 -1
- package/plugins/src/base/skills/lisa-queue-status/SKILL.md +10 -0
- package/plugins/src/base/skills/lisa-usage-accounting/SKILL.md +9 -0
- package/plugins/src/base/skills/lisa-usage-accounting/references/effectiveness.md +63 -0
|
@@ -1,21 +1,23 @@
|
|
|
1
1
|
# Tool Access Gate (load-bearing)
|
|
2
2
|
|
|
3
|
+
- If you can't reach something you need, such as a repository, a secret, an API, or a connector, say exactly what's missing in your first message and stop. Don't substitute, mock, or guess.
|
|
4
|
+
|
|
5
|
+
If the missing access is discovered after work begins, say exactly what's
|
|
6
|
+
missing in your next message and stop.
|
|
7
|
+
|
|
3
8
|
Only take on work you can finish. Before implementation starts — and again the
|
|
4
9
|
moment a new requirement surfaces mid-flow — enumerate every external tool the
|
|
5
10
|
work item needs (implementation, proof command, remote verification: AWS CLI,
|
|
6
11
|
Figma, Jam, Sentry, SonarCloud, PostHog, device/browser harnesses, databases,
|
|
7
12
|
deploy targets, …) and **prove access with a cheap read-only probe**, routing
|
|
8
|
-
through the matching `*-access` skill where one exists.
|
|
9
|
-
access
|
|
10
|
-
|
|
13
|
+
through the matching `*-access` skill where one exists. Resolve credentials through the documented access path before probing;
|
|
14
|
+
presence on PATH is not access. If that path or the probe cannot provide the
|
|
15
|
+
required access, report the gap and stop.
|
|
11
16
|
|
|
12
|
-
|
|
13
|
-
|
|
14
|
-
|
|
15
|
-
|
|
16
|
-
`blocked` verification verdict. Substituting weaker verification, mocking the
|
|
17
|
-
inaccessible system, guessing at tool contents, or narrowing scope to dodge the
|
|
18
|
-
gap are all forbidden.
|
|
17
|
+
The message must name the missing resource, the operation it blocks, and any
|
|
18
|
+
known access requirement, without exposing secret values. Substituting weaker
|
|
19
|
+
verification, mocking the inaccessible system, guessing at its contents, or
|
|
20
|
+
narrowing scope to dodge the gap are all forbidden.
|
|
19
21
|
|
|
20
|
-
Full contract (probe table,
|
|
22
|
+
Full contract (probe table, blocker report, resume path):
|
|
21
23
|
[reference/tool-access-gate.md](../reference/tool-access-gate.md).
|
|
@@ -1,9 +1,14 @@
|
|
|
1
1
|
# Tool Access Gate
|
|
2
2
|
|
|
3
|
+
- If you can't reach something you need, such as a repository, a secret, an API, or a connector, say exactly what's missing in your first message and stop. Don't substitute, mock, or guess.
|
|
4
|
+
|
|
5
|
+
If the missing access is discovered after work begins, say exactly what's
|
|
6
|
+
missing in your next message and stop.
|
|
7
|
+
|
|
3
8
|
A flow may only take on work it can actually finish. If completing a work item —
|
|
4
9
|
including its empirical verification — requires an external tool or system the
|
|
5
|
-
agent cannot access, the flow must **
|
|
6
|
-
|
|
10
|
+
agent cannot access, the flow must **tell the user exactly what access is missing and stop**, never
|
|
11
|
+
work around it. This is the flow-side arm of the factory
|
|
7
12
|
contract: intake validates that the factory has "the tooling *and provable
|
|
8
13
|
access to that tooling*"; this gate re-proves that promise at execution time and
|
|
9
14
|
enforces it for tools discovered mid-flow.
|
|
@@ -31,7 +36,8 @@ enforces it for tools discovered mid-flow.
|
|
|
31
36
|
environment.
|
|
32
37
|
2. **Continuously** — the moment a previously unknown tool requirement surfaces
|
|
33
38
|
mid-flow (e.g. verification turns out to need CloudWatch log capture), probe
|
|
34
|
-
it right then
|
|
39
|
+
it right then. If access is unavailable, report it and stop. Otherwise, record
|
|
40
|
+
the new tool + probe result in the same places the
|
|
35
41
|
preflight wrote to (the plan/tracker artifact and the affected tasks'
|
|
36
42
|
`metadata.required_access`) before continuing. Discovery timing changes
|
|
37
43
|
nothing about the protocol.
|
|
@@ -62,44 +68,31 @@ Example probes:
|
|
|
62
68
|
| Deploy target | reach the target environment with the credentials the verify step will use |
|
|
63
69
|
| Device/browser harness | the harness's own doctor/smoke entry (e.g. `playwright --version` plus a trivial headless launch) |
|
|
64
70
|
|
|
65
|
-
|
|
66
|
-
|
|
67
|
-
|
|
68
|
-
|
|
69
|
-
|
|
70
|
-
proceed.
|
|
71
|
+
Resolve credentials through the documented sources before probing: project
|
|
72
|
+
e2e config/fixtures, `.lisa.config.local.json` and environment variables, then
|
|
73
|
+
documented work-item credentials (e.g. `Sign-in Required`). If the documented
|
|
74
|
+
access path or probe cannot provide the required access, report the gap and
|
|
75
|
+
stop. Do not continue exploring substitute sources after confirming the gap.
|
|
71
76
|
|
|
72
|
-
Record
|
|
77
|
+
Record successful probes in the flow's plan/tracker artifact
|
|
73
78
|
(and task `metadata.required_access` where the flow's task contract carries
|
|
74
79
|
it), so the verifier can confirm the gate ran.
|
|
75
80
|
|
|
76
81
|
## On failure: break out, never work around
|
|
77
82
|
|
|
78
|
-
When
|
|
79
|
-
|
|
80
|
-
|
|
81
|
-
|
|
82
|
-
|
|
83
|
-
|
|
84
|
-
|
|
85
|
-
|
|
86
|
-
|
|
87
|
-
|
|
88
|
-
|
|
89
|
-
|
|
90
|
-
|
|
91
|
-
3. **Transition the work item to the configured blocked state** and apply the
|
|
92
|
-
configured `human_needed` / needs-human marker — missing access is a
|
|
93
|
-
**human-only blocker** (someone must provision credentials or grant access);
|
|
94
|
-
do not fabricate a build-ready ticket for it.
|
|
95
|
-
4. **Release the verification gate honestly**: write the verdict with
|
|
96
|
-
`status: "blocked"`, and mark each acceptance criterion whose proof depends
|
|
97
|
-
on the missing tool as `status: "blocked"` with the missing-access
|
|
98
|
-
diagnosis as its `evidence`; unaffected criteria keep their real
|
|
99
|
-
`pass`/`fail` result.
|
|
100
|
-
5. **Resume only when the probe passes.** When access is granted, re-run the
|
|
101
|
-
recorded probe before continuing; `repair-intake` re-validates blocked items
|
|
102
|
-
whose blockers cleared.
|
|
83
|
+
When required access is unavailable, tell the user and stop. Use the first
|
|
84
|
+
message if the gap is already known, or the next message if it is discovered
|
|
85
|
+
mid-task. The message must state:
|
|
86
|
+
|
|
87
|
+
- the exact resource that cannot be reached and the operation it blocks;
|
|
88
|
+
- the observed failure and, if known, the credential name, role, permission, or
|
|
89
|
+
invitation needed — never secret values, and never a guessed diagnosis;
|
|
90
|
+
- the read-only probe that must pass before work can resume.
|
|
91
|
+
|
|
92
|
+
Do not substitute, mock, guess, or continue other tasks as a workaround.
|
|
93
|
+
Reporting does not depend on access to a tracker, and it does not require
|
|
94
|
+
creating a new ticket. Resume after the required access is available and its
|
|
95
|
+
probe passes.
|
|
103
96
|
|
|
104
97
|
### Forbidden workarounds
|
|
105
98
|
|
|
@@ -116,6 +109,6 @@ None of the following ever substitutes for missing access:
|
|
|
116
109
|
ticket points at).
|
|
117
110
|
- Silently narrowing scope so the inaccessible part is "out of scope".
|
|
118
111
|
|
|
119
|
-
|
|
120
|
-
|
|
121
|
-
|
|
112
|
+
Keep results already obtained, but stop further work until the required access
|
|
113
|
+
is available. An inaccessible tracker does not prevent reporting the blocker
|
|
114
|
+
to the user. Do not claim blocked acceptance criteria have passed.
|
|
@@ -252,6 +252,10 @@ The rollup contract is additive across the hierarchy: PRDs may roll up Epics/Sto
|
|
|
252
252
|
- Re-running with the same logical entry set must produce byte-identical output. Rewriting a legacy
|
|
253
253
|
primary-only marker or the 2.222.0 transitional marker migrates once to the canonical
|
|
254
254
|
primary-plus-extension layout; subsequent rewrites are byte-identical.
|
|
255
|
-
- Do not
|
|
255
|
+
- Do not add timestamps to the section preamble or primary usage token. Sourced
|
|
256
|
+
lifecycle timestamps belong only in the optional effectiveness extension.
|
|
256
257
|
|
|
257
258
|
Idempotency is enforced by `entry_id` for direct entries and by the fixed rollup token field order for totals.
|
|
259
|
+
## Delivery observations
|
|
260
|
+
|
|
261
|
+
Read `references/effectiveness.md` in the `lisa-usage-accounting` skill for the optional effectiveness extension, clock definitions, and recurrence storage contract shared by every supported agent. The primary 17-field usage token remains unchanged.
|
|
@@ -260,11 +260,20 @@ computed **deterministically — never estimated by the model**:
|
|
|
260
260
|
|
|
261
261
|
1. **Normalize** the invariant text: trim leading/trailing whitespace,
|
|
262
262
|
collapse every internal whitespace run to a single space, lowercase.
|
|
263
|
-
2. **Hash**
|
|
263
|
+
2. **Hash** through the shared implementation: write the invariant to a text file
|
|
264
|
+
and run `lisa effectiveness fingerprint --input <file>`. This performs the
|
|
265
|
+
normalization above and returns the first 12 SHA-256 hex characters.
|
|
264
266
|
|
|
265
267
|
The same knowledge item therefore always produces the same key across runs,
|
|
266
268
|
regardless of which session computes it.
|
|
267
269
|
|
|
270
|
+
Read `lisa effectiveness report` for committed post-control recurrence counts by
|
|
271
|
+
the same `<surface>+<invariant-hash>` key. Cite the count and occurrence sources as
|
|
272
|
+
evidence when deciding whether a control helped. Missing history means unknown,
|
|
273
|
+
not a successful control. Counts do not automatically justify a new ticket: apply
|
|
274
|
+
the existing worth-doing judgment. The learning entry's content-version fingerprint
|
|
275
|
+
is a different identifier and must not be substituted for this invariant hash.
|
|
276
|
+
|
|
268
277
|
Before filing anything, search the tracker for the marker in
|
|
269
278
|
**open AND closed** issues — and key the search on the deterministic
|
|
270
279
|
`<surface>` prefix **first**, using the hash as disambiguation only, so even
|
|
@@ -69,6 +69,16 @@ For each inspected queue, report:
|
|
|
69
69
|
|
|
70
70
|
The report should stay terminal-first and immediately actionable: observable queue facts first, then the smallest useful next step.
|
|
71
71
|
|
|
72
|
+
## Delivery effort
|
|
73
|
+
|
|
74
|
+
Alongside queue counts, run the read-only `lisa effectiveness report`. Show observed
|
|
75
|
+
human interventions per accepted outcome, its numerator/denominator, and evidence
|
|
76
|
+
sources. State that this covers locally observed outcomes; absent reports or null
|
|
77
|
+
attention mean **unknown**, not zero. If queue-wide coverage has not been verified,
|
|
78
|
+
do not imply the ratio describes the whole queue. Show the separate run clocks and
|
|
79
|
+
committed recurring-failure counts when present. Do not create reports, change
|
|
80
|
+
tracker state, or use PR/token counts as outcome-value proxies from this skill.
|
|
81
|
+
|
|
72
82
|
## Pull request arming (#3903)
|
|
73
83
|
|
|
74
84
|
Alongside the two work queues, report the **arming state of the repo's open pull requests**. A PR whose `autoMergeRequest` is `null` is fully green and permanently unmergeable: every check passes, `mergeStateStatus` is green, nothing complains, and it waits forever. Green-and-unarmed and green-and-waiting read identically on every other surface, so this is the one place that asks the question.
|
|
@@ -140,6 +140,15 @@ the managed comment path.
|
|
|
140
140
|
|
|
141
141
|
### Step 3 — Apply the requested operation
|
|
142
142
|
|
|
143
|
+
For lifecycle rows, include the optional `effectiveness` observations defined in
|
|
144
|
+
[portable effectiveness reference](references/effectiveness.md): four separate clocks, explicit unknowns, sourced human
|
|
145
|
+
events, Lisa version, and worker configuration revision. Preserve extensions on old
|
|
146
|
+
rows. After the canonical host write passes readback, mirror the row through
|
|
147
|
+
`lisa effectiveness record --input <json-file>` for local status reports. On a
|
|
148
|
+
confirmed recurrence after its control shipped, use `lisa effectiveness recurrence
|
|
149
|
+
--input <json-file>` with the stable source event identity; commit that history
|
|
150
|
+
through normal checks. Do not infer human minutes or invent missing runtime telemetry.
|
|
151
|
+
|
|
143
152
|
All three operations use the shared utilities and rule contract; they differ only in which inputs
|
|
144
153
|
they require and whether they recompute child totals:
|
|
145
154
|
|
|
@@ -0,0 +1,63 @@
|
|
|
1
|
+
# Delivery observations (optional extension)
|
|
2
|
+
|
|
3
|
+
Keep the primary 17-field usage token unchanged. New writers attach
|
|
4
|
+
`LisaUsageEntry.effectiveness`; the shared serializer emits the adjacent
|
|
5
|
+
`lisa:usage-effectiveness` token correlated by `entry_id`. Older readers may ignore
|
|
6
|
+
the extension. New writers preserve it when rewriting historical rows.
|
|
7
|
+
|
|
8
|
+
Each observation has `schema: 1` and all these fields, using `null` for unknown:
|
|
9
|
+
|
|
10
|
+
- `lisaVersion`, `workerConfigRevision`: observed version/revision tags.
|
|
11
|
+
- `workerWallMs`: actual execution duration for this run.
|
|
12
|
+
- `feedbackMs`: measured blocking request-to-response samples in milliseconds.
|
|
13
|
+
Start when this worker issues the request; stop when its response is available.
|
|
14
|
+
Waiting on another agent is queue latency and is excluded. `[]` means a measured
|
|
15
|
+
run with no such requests; `null` means unavailable telemetry.
|
|
16
|
+
- `workerSource`: runtime evidence identifier for wall time and feedback samples.
|
|
17
|
+
- `attention`: observed `{id, source}` human interventions. Derive these from
|
|
18
|
+
non-bot ready-role flips, blocked-item answers, and review interventions. Dedupe
|
|
19
|
+
by stable tracker event identity. `[]` means the relevant history was inspected
|
|
20
|
+
and contained no interventions; `null` means it was not measured. Never infer
|
|
21
|
+
human minutes from gaps between comments.
|
|
22
|
+
- `readyAt`, `acceptedAt`: canonical UTC timestamps from the configured ready-role
|
|
23
|
+
flip and terminal accepted outcome. The elapsed clock excludes pre-ready intake.
|
|
24
|
+
A merge or queued deployment alone is not acceptance. `lifecycleSources` holds
|
|
25
|
+
the tracker/release evidence for these timestamps.
|
|
26
|
+
- `reworkSources`: observed reclaims, QA bounces, or reverts. The existing
|
|
27
|
+
`lisa-delivery-effectiveness` skill owns rework rates and other outcome measures;
|
|
28
|
+
these are observations, not a competing score.
|
|
29
|
+
|
|
30
|
+
At the existing implement, verify, and intake accounting milestones, supply these
|
|
31
|
+
observations even if every measurement is unknown. Collect runtime observations
|
|
32
|
+
as work proceeds; do not reconstruct feedback samples from memory at the end.
|
|
33
|
+
Runtime APIs differ across agents: unavailable measurements remain null on any
|
|
34
|
+
agent rather than being estimated from tool/PR/token counts.
|
|
35
|
+
|
|
36
|
+
After the canonical usage row is read back, mirror `{entryId, artifactRef,
|
|
37
|
+
effectiveness}` with `lisa effectiveness record --input <json-file>`. Mirrors under
|
|
38
|
+
`.lisa/effectiveness/` are ignored and disposable; reconstruct them from usage
|
|
39
|
+
extensions when needed. Do not commit timing reports or store secrets in sources.
|
|
40
|
+
The command refuses a non-ignored destination. Report write failures explicitly;
|
|
41
|
+
they must not erase the canonical accounting row or claim measurement succeeded.
|
|
42
|
+
|
|
43
|
+
Record an actual post-control failure using `lisa effectiveness recurrence --input
|
|
44
|
+
<json-file>`. The object contains `invariant`, `surface`, `controlRef`,
|
|
45
|
+
`controlShippedAt`, `occurrenceRef`, and `occurredAt`. Both timestamps use canonical
|
|
46
|
+
UTC ISO strings with milliseconds. Verify that the control/learning really shipped
|
|
47
|
+
before this occurrence; initial failures and unshipped proposals are not recurrences.
|
|
48
|
+
Use a stable occurrence source, such as a test-run failure or tracker event, so
|
|
49
|
+
replays and reports from different agents identify the same event.
|
|
50
|
+
|
|
51
|
+
`.lisa/RECURRENCES.jsonl` is committed history: distinct occurrence identities are
|
|
52
|
+
the durable count. Its existing built-in `merge=union` attribute retains concurrent
|
|
53
|
+
branch appends, and the reader deduplicates replayed events. Never hand-increment
|
|
54
|
+
a counter, compact away occurrences, or add counts to the bounded learnings ledger.
|
|
55
|
+
Normal review, secret checks, and push requirements still apply to this file.
|
|
56
|
+
The writer refuses missing merge setup and conflicting evidence for one identity.
|
|
57
|
+
|
|
58
|
+
`lisa effectiveness report` is read-only. Its per-run feedback distribution, wall
|
|
59
|
+
time, attention count, and ready-to-accepted duration remain separate. Attention
|
|
60
|
+
ratios cover only the locally observed accepted outcomes, with the reported sources
|
|
61
|
+
and denominator. Missing local records are unknown coverage, not zero factory effort.
|
|
62
|
+
Rebuild local records for the requested reporting scope before making a scope-wide
|
|
63
|
+
claim; never extrapolate a local subset to the whole queue.
|