zelari-code 2.1.0 → 2.1.2
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/README.md +4 -4
- package/package.json +8 -5
- package/dist/cli/ast/engine.test.js +0 -101
- package/dist/cli/ast/engine.test.js.map +0 -1
- package/dist/cli/ast/tools.test.js +0 -124
- package/dist/cli/ast/tools.test.js.map +0 -1
- package/dist/cli/headless.test.js +0 -107
- package/dist/cli/headless.test.js.map +0 -1
- package/dist/cli/headlessE2eSession.test.js +0 -189
- package/dist/cli/headlessE2eSession.test.js.map +0 -1
- package/dist/cli/headlessHistorySeed.test.js +0 -158
- package/dist/cli/headlessHistorySeed.test.js.map +0 -1
- package/dist/cli/headlessSessionEvent.test.js +0 -83
- package/dist/cli/headlessSessionEvent.test.js.map +0 -1
- package/dist/cli/headlessSpine.test.js +0 -153
- package/dist/cli/headlessSpine.test.js.map +0 -1
- package/dist/cli/hooks/observationStore.test.js +0 -135
- package/dist/cli/hooks/observationStore.test.js.map +0 -1
- package/dist/cli/hooks/sessionSurface.test.js +0 -112
- package/dist/cli/hooks/sessionSurface.test.js.map +0 -1
- package/dist/cli/kraken/candidateRegistry.test.js +0 -211
- package/dist/cli/kraken/candidateRegistry.test.js.map +0 -1
- package/dist/cli/kraken/completionGate.test.js +0 -172
- package/dist/cli/kraken/completionGate.test.js.map +0 -1
- package/dist/cli/kraken/evidenceAnchoring.test.js +0 -92
- package/dist/cli/kraken/evidenceAnchoring.test.js.map +0 -1
- package/dist/cli/kraken/evidenceProvenance.test.js +0 -113
- package/dist/cli/kraken/evidenceProvenance.test.js.map +0 -1
- package/dist/cli/kraken/executor.workbench.test.js +0 -134
- package/dist/cli/kraken/executor.workbench.test.js.map +0 -1
- package/dist/cli/kraken/metrics.test.js +0 -145
- package/dist/cli/kraken/metrics.test.js.map +0 -1
- package/dist/cli/kraken/nativeVerification.test.js +0 -207
- package/dist/cli/kraken/nativeVerification.test.js.map +0 -1
- package/dist/cli/kraken/planner.test.js +0 -43
- package/dist/cli/kraken/planner.test.js.map +0 -1
- package/dist/cli/kraken/scriptPlanner.test.js +0 -152
- package/dist/cli/kraken/scriptPlanner.test.js.map +0 -1
- package/dist/cli/kraken/selectionPlaybook.test.js +0 -55
- package/dist/cli/kraken/selectionPlaybook.test.js.map +0 -1
- package/dist/cli/kraken/skillSuggest.test.js +0 -157
- package/dist/cli/kraken/skillSuggest.test.js.map +0 -1
- package/dist/cli/kraken/strictDefaults.test.js +0 -149
- package/dist/cli/kraken/strictDefaults.test.js.map +0 -1
- package/dist/cli/kraken/strictGatePackIndependence.test.js +0 -112
- package/dist/cli/kraken/strictGatePackIndependence.test.js.map +0 -1
- package/dist/cli/kraken/turnRuntime.test.js +0 -377
- package/dist/cli/kraken/turnRuntime.test.js.map +0 -1
- package/dist/cli/kraken/verificationBridge.session.test.js +0 -104
- package/dist/cli/kraken/verificationBridge.session.test.js.map +0 -1
- package/dist/cli/kraken/verificationBridge.test.js +0 -177
- package/dist/cli/kraken/verificationBridge.test.js.map +0 -1
- package/dist/cli/kraken/verifier.test.js +0 -232
- package/dist/cli/kraken/verifier.test.js.map +0 -1
- package/dist/cli/kraken/verifierAdvisoryLock.test.js +0 -154
- package/dist/cli/kraken/verifierAdvisoryLock.test.js.map +0 -1
- package/dist/cli/kraken/verifierLifecycle.test.js +0 -181
- package/dist/cli/kraken/verifierLifecycle.test.js.map +0 -1
- package/dist/cli/kraken/verifierRoundTrip.test.js +0 -166
- package/dist/cli/kraken/verifierRoundTrip.test.js.map +0 -1
- package/dist/cli/kraken/verifierSettings.test.js +0 -237
- package/dist/cli/kraken/verifierSettings.test.js.map +0 -1
- package/dist/cli/kraken/verifyReport.test.js +0 -111
- package/dist/cli/kraken/verifyReport.test.js.map +0 -1
- package/dist/cli/kraken/weaknessMeter.test.js +0 -212
- package/dist/cli/kraken/weaknessMeter.test.js.map +0 -1
- package/dist/cli/kraken/workbench.test.js +0 -253
- package/dist/cli/kraken/workbench.test.js.map +0 -1
- package/dist/cli/kraken/workbenchView.test.js +0 -130
- package/dist/cli/kraken/workbenchView.test.js.map +0 -1
- package/dist/cli/legacyContextIsolation.test.js +0 -65
- package/dist/cli/legacyContextIsolation.test.js.map +0 -1
- package/dist/cli/lsp/tools.test.js +0 -90
- package/dist/cli/lsp/tools.test.js.map +0 -1
- package/dist/cli/missionContinuation.test.js +0 -117
- package/dist/cli/missionContinuation.test.js.map +0 -1
- package/dist/cli/profileMatrix.test.js +0 -130
- package/dist/cli/profileMatrix.test.js.map +0 -1
- package/dist/cli/provider/capabilities.test.js +0 -46
- package/dist/cli/provider/capabilities.test.js.map +0 -1
- package/dist/cli/provider/openai-compatible.timeout.test.js +0 -72
- package/dist/cli/provider/openai-compatible.timeout.test.js.map +0 -1
- package/dist/cli/providerConfig.test.js +0 -85
- package/dist/cli/providerConfig.test.js.map +0 -1
- package/dist/cli/sessionReplayInvariant.test.js +0 -187
- package/dist/cli/sessionReplayInvariant.test.js.map +0 -1
- package/dist/cli/sessionSpine.test.js +0 -171
- package/dist/cli/sessionSpine.test.js.map +0 -1
- package/dist/cli/thinking.test.js +0 -214
- package/dist/cli/thinking.test.js.map +0 -1
- package/dist/cli/tools/inspectCommand.schema.test.js +0 -82
- package/dist/cli/tools/inspectCommand.schema.test.js.map +0 -1
- package/dist/cli/tools/inspectCommand.test.js +0 -311
- package/dist/cli/tools/inspectCommand.test.js.map +0 -1
- package/dist/cli/tools/krakenCsvFanout.test.js +0 -200
- package/dist/cli/tools/krakenCsvFanout.test.js.map +0 -1
- package/dist/cli/tools/krakenSelectTool.test.js +0 -246
- package/dist/cli/tools/krakenSelectTool.test.js.map +0 -1
- package/dist/cli/tools/observeBatch.test.js +0 -329
- package/dist/cli/tools/observeBatch.test.js.map +0 -1
- package/dist/cli/tools/taskTool.dynamicChecks.test.js +0 -156
- package/dist/cli/tools/taskTool.dynamicChecks.test.js.map +0 -1
- package/dist/cli/tools/taskTool.planSafety.test.js +0 -280
- package/dist/cli/tools/taskTool.planSafety.test.js.map +0 -1
- package/dist/cli/tools/taskTool.verifyReport.test.js +0 -140
- package/dist/cli/tools/taskTool.verifyReport.test.js.map +0 -1
- package/dist/cli/utils/contextGrowthSummary.test.js +0 -74
- package/dist/cli/utils/contextGrowthSummary.test.js.map +0 -1
- package/dist/cli/workspace/planDriftCheck.test.js +0 -146
- package/dist/cli/workspace/planDriftCheck.test.js.map +0 -1
- package/dist/cli/workspace/planStore.test.js +0 -295
- package/dist/cli/workspace/planStore.test.js.map +0 -1
- package/dist/cli/zelariMission.test.js +0 -44
- package/dist/cli/zelariMission.test.js.map +0 -1
|
@@ -1,280 +0,0 @@
|
|
|
1
|
-
/**
|
|
2
|
-
* taskTool.planSafety.test — Fase 1 / ADR-0020 (Kraken Verified Selection).
|
|
3
|
-
*
|
|
4
|
-
* DoD della Fase 1:
|
|
5
|
-
* - PLAN può spawnare explore (task registrato, policy explore-only)
|
|
6
|
-
* - PLAN non può spawnare general né verify
|
|
7
|
-
* - PLAN non espone write/bash (read-only invariato)
|
|
8
|
-
* - BUILD task continua a comportarsi come prima
|
|
9
|
-
* - Cancel parent cancella il tentacolo (signal propagato a runTentacle)
|
|
10
|
-
*/
|
|
11
|
-
import { describe, expect, it, beforeEach } from 'vitest';
|
|
12
|
-
import { tmpdir } from 'node:os';
|
|
13
|
-
import path from 'node:path';
|
|
14
|
-
import fs from 'node:fs';
|
|
15
|
-
import { fileURLToPath } from 'node:url';
|
|
16
|
-
import { ToolRegistry } from '@zelari/core/harness/tools/registry';
|
|
17
|
-
import { createBuiltinToolRegistry } from '../toolRegistry.js';
|
|
18
|
-
import { createTaskTool, resetTaskSpawnCount, } from './taskTool.js';
|
|
19
|
-
import { isKrakenSelectionEnabled, krakenCandidates, resetKrakenCandidates, } from '../kraken/candidateRegistry.js';
|
|
20
|
-
import { AuditLogger } from '../safety/auditLogger.js';
|
|
21
|
-
import { defaultPermissionPolicy } from '../safety/toolPermissions.js';
|
|
22
|
-
const repoRoot = path.resolve(path.dirname(fileURLToPath(import.meta.url)), '..', '..', '..');
|
|
23
|
-
function tmpRoot() {
|
|
24
|
-
return fs.mkdtempSync(path.join(tmpdir(), 'zelari-plan-safe-'));
|
|
25
|
-
}
|
|
26
|
-
function makeRegistry(extra = {}) {
|
|
27
|
-
return createBuiltinToolRegistry({
|
|
28
|
-
root: repoRoot,
|
|
29
|
-
lspProvider: null,
|
|
30
|
-
audit: new AuditLogger(path.join(tmpdir(), `zelari-plansafe-${Date.now()}-${Math.random().toString(36).slice(2)}.log`)),
|
|
31
|
-
permissionPolicy: defaultPermissionPolicy({ auto: true }),
|
|
32
|
-
...extra,
|
|
33
|
-
});
|
|
34
|
-
}
|
|
35
|
-
function makeCtx(cwd = tmpRoot()) {
|
|
36
|
-
return {
|
|
37
|
-
signal: new AbortController().signal,
|
|
38
|
-
cwd,
|
|
39
|
-
audit: () => undefined,
|
|
40
|
-
sessionId: 'plan-safety-test',
|
|
41
|
-
};
|
|
42
|
-
}
|
|
43
|
-
/** Deps fake: null context = "no provider" path, enough to observe the gate. */
|
|
44
|
-
function probeDeps() {
|
|
45
|
-
const calls = [];
|
|
46
|
-
const deps = {
|
|
47
|
-
createSubAgentContext: async ({ agent }) => {
|
|
48
|
-
calls.push(agent);
|
|
49
|
-
return null;
|
|
50
|
-
},
|
|
51
|
-
allowWorktree: false,
|
|
52
|
-
};
|
|
53
|
-
return { deps, calls };
|
|
54
|
-
}
|
|
55
|
-
/** Deps fake with a valid context + scripted harness (signal tests). */
|
|
56
|
-
function scriptedDeps() {
|
|
57
|
-
let started = 0;
|
|
58
|
-
const deps = {
|
|
59
|
-
createSubAgentContext: async ({ agent }) => {
|
|
60
|
-
const ctx = {
|
|
61
|
-
providerStream: (() => {
|
|
62
|
-
throw new Error('not invoked by the scripted harness');
|
|
63
|
-
}),
|
|
64
|
-
model: 'test-model',
|
|
65
|
-
provider: 'test-provider',
|
|
66
|
-
registry: new ToolRegistry(),
|
|
67
|
-
tools: [],
|
|
68
|
-
agent,
|
|
69
|
-
};
|
|
70
|
-
return ctx;
|
|
71
|
-
},
|
|
72
|
-
harnessFactory: () => ({
|
|
73
|
-
run: async function* () {
|
|
74
|
-
started++;
|
|
75
|
-
yield { type: 'message_start' };
|
|
76
|
-
yield { type: 'message_delta', delta: 'tentacle conclusion' };
|
|
77
|
-
yield { type: 'message_end' };
|
|
78
|
-
},
|
|
79
|
-
}),
|
|
80
|
-
allowWorktree: false,
|
|
81
|
-
};
|
|
82
|
-
return { deps, harnessStarted: () => started };
|
|
83
|
-
}
|
|
84
|
-
describe('registry gating (ADR-0020 Fase 1)', () => {
|
|
85
|
-
it('planMode registry registers task AND stays read-only', () => {
|
|
86
|
-
const { tools } = makeRegistry({ planMode: true });
|
|
87
|
-
const names = tools.map((t) => t.name);
|
|
88
|
-
expect(names).toContain('task');
|
|
89
|
-
expect(names).not.toContain('write_file');
|
|
90
|
-
expect(names).not.toContain('edit_file');
|
|
91
|
-
expect(names).not.toContain('apply_diff');
|
|
92
|
-
expect(names).not.toContain('bash');
|
|
93
|
-
});
|
|
94
|
-
it('planMode task is wired explore-only (restricted schema + description)', () => {
|
|
95
|
-
const { registry } = makeRegistry({ planMode: true });
|
|
96
|
-
const task = registry.get('task');
|
|
97
|
-
expect(task).toBeTruthy();
|
|
98
|
-
expect(task?.description).toContain('RESTRICTED in this mode');
|
|
99
|
-
// zod enum narrowed: general is not a valid value anymore
|
|
100
|
-
const bad = task?.inputSchema
|
|
101
|
-
?.safeParse({ description: 'd', prompt: 'p', agent: 'general' });
|
|
102
|
-
expect(bad?.success).toBe(false);
|
|
103
|
-
const good = task?.inputSchema
|
|
104
|
-
?.safeParse({ description: 'd', prompt: 'p', agent: 'explore' });
|
|
105
|
-
expect(good?.success).toBe(true);
|
|
106
|
-
});
|
|
107
|
-
it('planExploreTask=false restores the pre-ADR-0020 behaviour (no task in plan)', () => {
|
|
108
|
-
const { tools } = makeRegistry({ planMode: true, planExploreTask: false });
|
|
109
|
-
expect(tools.map((t) => t.name)).not.toContain('task');
|
|
110
|
-
});
|
|
111
|
-
it('full BUILD registry keeps task unrestricted (regression)', () => {
|
|
112
|
-
const { registry, tools } = makeRegistry();
|
|
113
|
-
expect(tools.map((t) => t.name)).toContain('task');
|
|
114
|
-
expect(registry.get('task')?.description).not.toContain('RESTRICTED in this mode');
|
|
115
|
-
});
|
|
116
|
-
it('explicit readOnly registry omits task (sub-agent anti-recursion, regression)', () => {
|
|
117
|
-
const { tools } = makeRegistry({ readOnly: true });
|
|
118
|
-
expect(tools.map((t) => t.name)).not.toContain('task');
|
|
119
|
-
});
|
|
120
|
-
it('explore sub-profile omits task (no nesting, regression)', () => {
|
|
121
|
-
const { tools } = makeRegistry({ profile: 'explore' });
|
|
122
|
-
expect(tools.map((t) => t.name)).not.toContain('task');
|
|
123
|
-
});
|
|
124
|
-
});
|
|
125
|
-
describe('task policy (ADR-0020 Fase 1)', () => {
|
|
126
|
-
beforeEach(() => {
|
|
127
|
-
resetTaskSpawnCount();
|
|
128
|
-
});
|
|
129
|
-
it('explore-only policy rejects agent=general BEFORE spawning or consuming budget', async () => {
|
|
130
|
-
const { deps, calls } = probeDeps();
|
|
131
|
-
const tool = createTaskTool(deps, { allowedAgents: ['explore'] });
|
|
132
|
-
const res = await tool.execute({ description: 'd', prompt: 'p', agent: 'general' }, makeCtx());
|
|
133
|
-
expect(res.ok).toBe(false);
|
|
134
|
-
if (!res.ok) {
|
|
135
|
-
expect(res.error).toMatch(/not allowed in this mode/i);
|
|
136
|
-
expect(res.error).toContain('explore');
|
|
137
|
-
}
|
|
138
|
-
expect(calls).toHaveLength(0);
|
|
139
|
-
// budget NOT consumed: the very next spawn is still allowed
|
|
140
|
-
const next = await tool.execute({ description: 'd', prompt: 'p', agent: 'verify' }, makeCtx());
|
|
141
|
-
expect(next.ok).toBe(false);
|
|
142
|
-
if (!next.ok)
|
|
143
|
-
expect(next.error).toMatch(/not allowed in this mode/i);
|
|
144
|
-
expect(calls).toHaveLength(0);
|
|
145
|
-
});
|
|
146
|
-
it('explore-only policy accepts the default agent (explore reaches the context factory)', async () => {
|
|
147
|
-
const { deps, calls } = probeDeps();
|
|
148
|
-
const tool = createTaskTool(deps, { allowedAgents: ['explore'] });
|
|
149
|
-
const res = await tool.execute({ description: 'd', prompt: 'p' }, makeCtx());
|
|
150
|
-
expect(calls).toEqual(['explore']);
|
|
151
|
-
// fake context is null → "no provider", which PROVES we got past the gate
|
|
152
|
-
expect(res.ok).toBe(false);
|
|
153
|
-
if (!res.ok)
|
|
154
|
-
expect(res.error).toMatch(/no provider/);
|
|
155
|
-
});
|
|
156
|
-
it('unrestricted BUILD tool still accepts general (regression)', async () => {
|
|
157
|
-
const { deps, calls } = probeDeps();
|
|
158
|
-
const tool = createTaskTool(deps);
|
|
159
|
-
const res = await tool.execute({ description: 'd', prompt: 'p', agent: 'general' }, makeCtx());
|
|
160
|
-
expect(calls).toEqual(['general']);
|
|
161
|
-
expect(res.ok).toBe(false);
|
|
162
|
-
if (!res.ok)
|
|
163
|
-
expect(res.error).toMatch(/no provider/);
|
|
164
|
-
});
|
|
165
|
-
});
|
|
166
|
-
describe('cancellation propagation (ADR-0020 Fase 1)', () => {
|
|
167
|
-
beforeEach(() => {
|
|
168
|
-
resetTaskSpawnCount();
|
|
169
|
-
});
|
|
170
|
-
it('an already-aborted parent signal cancels the tentacle instead of running it', async () => {
|
|
171
|
-
const { deps, harnessStarted } = scriptedDeps();
|
|
172
|
-
const tool = createTaskTool(deps);
|
|
173
|
-
const controller = new AbortController();
|
|
174
|
-
controller.abort();
|
|
175
|
-
const res = await tool.execute({ description: 'd', prompt: 'p', agent: 'explore' }, { ...makeCtx(), signal: controller.signal });
|
|
176
|
-
expect(res.ok).toBe(false);
|
|
177
|
-
if (!res.ok)
|
|
178
|
-
expect(res.error).toMatch(/cancelled/i);
|
|
179
|
-
expect(harnessStarted()).toBe(0);
|
|
180
|
-
});
|
|
181
|
-
it('a live signal does not break the normal tentacle run (regression)', async () => {
|
|
182
|
-
const { deps, harnessStarted } = scriptedDeps();
|
|
183
|
-
const tool = createTaskTool(deps);
|
|
184
|
-
const res = await tool.execute({ description: 'd', prompt: 'p', agent: 'explore' }, makeCtx());
|
|
185
|
-
expect(res.ok).toBe(true);
|
|
186
|
-
if (res.ok)
|
|
187
|
-
expect(res.value.result).toContain('tentacle conclusion');
|
|
188
|
-
expect(harnessStarted()).toBe(1);
|
|
189
|
-
});
|
|
190
|
-
});
|
|
191
|
-
describe('candidate contract (ADR-0020 Fase 3)', () => {
|
|
192
|
-
const ctx = makeCtx();
|
|
193
|
-
beforeEach(() => {
|
|
194
|
-
resetTaskSpawnCount();
|
|
195
|
-
resetKrakenCandidates();
|
|
196
|
-
});
|
|
197
|
-
it('purpose=candidate is refused while the alpha flag is OFF', async () => {
|
|
198
|
-
const prev = process.env.ZELARI_KRAKEN_SELECTION;
|
|
199
|
-
delete process.env.ZELARI_KRAKEN_SELECTION;
|
|
200
|
-
try {
|
|
201
|
-
const { deps, calls } = probeDeps();
|
|
202
|
-
const tool = createTaskTool(deps);
|
|
203
|
-
const res = await tool.execute({ description: 'd', prompt: 'p', purpose: 'candidate' }, ctx);
|
|
204
|
-
expect(res.ok).toBe(false);
|
|
205
|
-
if (!res.ok)
|
|
206
|
-
expect(res.error).toMatch(/ZELARI_KRAKEN_SELECTION=1/);
|
|
207
|
-
expect(calls).toHaveLength(0);
|
|
208
|
-
expect(krakenCandidates()).toHaveLength(0);
|
|
209
|
-
}
|
|
210
|
-
finally {
|
|
211
|
-
if (prev === undefined)
|
|
212
|
-
delete process.env.ZELARI_KRAKEN_SELECTION;
|
|
213
|
-
else
|
|
214
|
-
process.env.ZELARI_KRAKEN_SELECTION = prev;
|
|
215
|
-
}
|
|
216
|
-
});
|
|
217
|
-
it('candidate forces agent=explore (general rejected without spawning)', async () => {
|
|
218
|
-
process.env.ZELARI_KRAKEN_SELECTION = '1';
|
|
219
|
-
try {
|
|
220
|
-
const { deps, calls } = probeDeps();
|
|
221
|
-
const tool = createTaskTool(deps);
|
|
222
|
-
const res = await tool.execute({ description: 'd', prompt: 'p', agent: 'general', purpose: 'candidate' }, ctx);
|
|
223
|
-
expect(res.ok).toBe(false);
|
|
224
|
-
if (!res.ok)
|
|
225
|
-
expect(res.error).toMatch(/forces agent=explore/);
|
|
226
|
-
expect(calls).toHaveLength(0);
|
|
227
|
-
expect(isKrakenSelectionEnabled()).toBe(true);
|
|
228
|
-
}
|
|
229
|
-
finally {
|
|
230
|
-
delete process.env.ZELARI_KRAKEN_SELECTION;
|
|
231
|
-
}
|
|
232
|
-
});
|
|
233
|
-
it('candidate failures are registered as malformed (slot consumed, evidence preserved)', async () => {
|
|
234
|
-
process.env.ZELARI_KRAKEN_SELECTION = '1';
|
|
235
|
-
try {
|
|
236
|
-
// probeDeps returns a null context → tentacle fails with no provider
|
|
237
|
-
const { deps } = probeDeps();
|
|
238
|
-
const tool = createTaskTool(deps);
|
|
239
|
-
const res = await tool.execute({ description: 'd', prompt: 'p', purpose: 'candidate' }, ctx);
|
|
240
|
-
expect(res.ok).toBe(false);
|
|
241
|
-
const all = krakenCandidates();
|
|
242
|
-
expect(all).toHaveLength(1);
|
|
243
|
-
expect(all[0].status).toBe('malformed');
|
|
244
|
-
}
|
|
245
|
-
finally {
|
|
246
|
-
delete process.env.ZELARI_KRAKEN_SELECTION;
|
|
247
|
-
}
|
|
248
|
-
});
|
|
249
|
-
});
|
|
250
|
-
describe('kraken_select gating (ADR-0020 Fase 4)', () => {
|
|
251
|
-
it('absent by default (flag off → registry unchanged)', () => {
|
|
252
|
-
const { tools } = makeRegistry();
|
|
253
|
-
expect(tools.find((t) => t.name === 'kraken_select')).toBeUndefined();
|
|
254
|
-
});
|
|
255
|
-
it('registered on the parent full registry when krakenSelect=true', () => {
|
|
256
|
-
const { tools, registry } = makeRegistry({ krakenSelect: true });
|
|
257
|
-
const summary = tools.find((t) => t.name === 'kraken_select');
|
|
258
|
-
expect(summary).toBeDefined();
|
|
259
|
-
expect(summary?.permissions).toContain('read');
|
|
260
|
-
expect(registry.get('kraken_select')).toBeDefined();
|
|
261
|
-
});
|
|
262
|
-
it('stays available in PLAN mode (selection is plan-safe)', () => {
|
|
263
|
-
const { tools } = makeRegistry({ krakenSelect: true, planMode: true });
|
|
264
|
-
expect(tools.find((t) => t.name === 'kraken_select')).toBeDefined();
|
|
265
|
-
});
|
|
266
|
-
it('NEVER registered on read-only sub-agent profiles (tentacles)', () => {
|
|
267
|
-
const { tools, registry } = makeRegistry({ krakenSelect: true, readOnly: true });
|
|
268
|
-
expect(tools.find((t) => t.name === 'kraken_select')).toBeUndefined();
|
|
269
|
-
expect(registry.get('kraken_select')).toBeUndefined();
|
|
270
|
-
});
|
|
271
|
-
it('execute is safe without deps wiring (no candidates → typedOk no-op)', async () => {
|
|
272
|
-
resetKrakenCandidates();
|
|
273
|
-
const { registry } = makeRegistry({ krakenSelect: true });
|
|
274
|
-
const tool = registry.get('kraken_select');
|
|
275
|
-
expect(tool).toBeDefined();
|
|
276
|
-
const res = await tool.execute({}, makeCtx());
|
|
277
|
-
expect(res.ok).toBe(true);
|
|
278
|
-
});
|
|
279
|
-
});
|
|
280
|
-
//# sourceMappingURL=taskTool.planSafety.test.js.map
|
|
@@ -1 +0,0 @@
|
|
|
1
|
-
{"version":3,"file":"taskTool.planSafety.test.js","sourceRoot":"","sources":["../../../src/cli/tools/taskTool.planSafety.test.ts"],"names":[],"mappings":"AAAA;;;;;;;;;GASG;AACH,OAAO,EAAE,QAAQ,EAAE,MAAM,EAAE,EAAE,EAAE,UAAU,EAAE,MAAM,QAAQ,CAAC;AAC1D,OAAO,EAAE,MAAM,EAAE,MAAM,SAAS,CAAC;AACjC,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,MAAM,SAAS,CAAC;AACzB,OAAO,EAAE,aAAa,EAAE,MAAM,UAAU,CAAC;AAEzC,OAAO,EAAE,YAAY,EAAE,MAAM,qCAAqC,CAAC;AAEnE,OAAO,EAAE,yBAAyB,EAAE,MAAM,oBAAoB,CAAC;AAC/D,OAAO,EACL,cAAc,EACd,mBAAmB,GAGpB,MAAM,eAAe,CAAC;AACvB,OAAO,EACL,wBAAwB,EACxB,gBAAgB,EAChB,qBAAqB,GACtB,MAAM,gCAAgC,CAAC;AACxC,OAAO,EAAE,WAAW,EAAE,MAAM,0BAA0B,CAAC;AACvD,OAAO,EAAE,uBAAuB,EAAE,MAAM,8BAA8B,CAAC;AAEvE,MAAM,QAAQ,GAAG,IAAI,CAAC,OAAO,CAC3B,IAAI,CAAC,OAAO,CAAC,aAAa,CAAC,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,EAC5C,IAAI,EACJ,IAAI,EACJ,IAAI,CACL,CAAC;AAEF,SAAS,OAAO;IACd,OAAO,EAAE,CAAC,WAAW,CAAC,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,EAAE,mBAAmB,CAAC,CAAC,CAAC;AAClE,CAAC;AAED,SAAS,YAAY,CAAC,QAAiC,EAAE;IACvD,OAAO,yBAAyB,CAAC;QAC/B,IAAI,EAAE,QAAQ;QACd,WAAW,EAAE,IAAI;QACjB,KAAK,EAAE,IAAI,WAAW,CACpB,IAAI,CAAC,IAAI,CACP,MAAM,EAAE,EACR,mBAAmB,IAAI,CAAC,GAAG,EAAE,IAAI,IAAI,CAAC,MAAM,EAAE,CAAC,QAAQ,CAAC,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAC3E,CACF;QACD,gBAAgB,EAAE,uBAAuB,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,CAAC;QACzD,GAAG,KAAK;KACT,CAAC,CAAC;AACL,CAAC;AAED,SAAS,OAAO,CAAC,GAAG,GAAG,OAAO,EAAE;IAC9B,OAAO;QACL,MAAM,EAAE,IAAI,eAAe,EAAE,CAAC,MAAM;QACpC,GAAG;QACH,KAAK,EAAE,GAAG,EAAE,CAAC,SAAS;QACtB,SAAS,EAAE,kBAAkB;KAC9B,CAAC;AACJ,CAAC;AAED,gFAAgF;AAChF,SAAS,SAAS;IAChB,MAAM,KAAK,GAAa,EAAE,CAAC;IAC3B,MAAM,IAAI,GAAiB;QACzB,qBAAqB,EAAE,KAAK,EAAE,EAAE,KAAK,EAAE,EAAE,EAAE;YACzC,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;YAClB,OAAO,IAAI,CAAC;QACd,CAAC;QACD,aAAa,EAAE,KAAK;KACrB,CAAC;IACF,OAAO,EAAE,IAAI,EAAE,KAAK,EAAE,CAAC;AACzB,CAAC;AAED,wEAAwE;AACxE,SAAS,YAAY;IACnB,IAAI,OAAO,GAAG,CAAC,CAAC;IAChB,MAAM,IAAI,GAAiB;QACzB,qBAAqB,EAAE,KAAK,EAAE,EAAE,KAAK,EAAE,EAAE,EAAE;YACzC,MAAM,GAAG,GAAoB;gBAC3B,cAAc,EAAE,CAAC,GAAG,EAAE;oBACpB,MAAM,IAAI,KAAK,CAAC,qCAAqC,CAAC,CAAC;gBACzD,CAAC,CAAiD;gBAClD,KAAK,EAAE,YAAY;gBACnB,QAAQ,EAAE,eAAe;gBACzB,QAAQ,EAAE,IAAI,YAAY,EAAE;gBAC5B,KAAK,EAAE,EAAE;gBACT,KAAK;aACN,CAAC;YACF,OAAO,GAAG,CAAC;QACb,CAAC;QACD,cAAc,EAAE,GAAG,EAAE,CAAC,CAAC;YACrB,GAAG,EAAE,KAAK,SAAS,CAAC;gBAClB,OAAO,EAAE,CAAC;gBACV,MAAM,EAAE,IAAI,EAAE,eAAe,EAAgB,CAAC;gBAC9C,MAAM,EAAE,IAAI,EAAE,eAAe,EAAE,KAAK,EAAE,qBAAqB,EAAgB,CAAC;gBAC5E,MAAM,EAAE,IAAI,EAAE,aAAa,EAAgB,CAAC;YAC9C,CAAC;SACF,CAAC;QACF,aAAa,EAAE,KAAK;KACrB,CAAC;IACF,OAAO,EAAE,IAAI,EAAE,cAAc,EAAE,GAAG,EAAE,CAAC,OAAO,EAAE,CAAC;AACjD,CAAC;AAED,QAAQ,CAAC,mCAAmC,EAAE,GAAG,EAAE;IACjD,EAAE,CAAC,sDAAsD,EAAE,GAAG,EAAE;QAC9D,MAAM,EAAE,KAAK,EAAE,GAAG,YAAY,CAAC,EAAE,QAAQ,EAAE,IAAI,EAAE,CAAC,CAAC;QACnD,MAAM,KAAK,GAAG,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;QACvC,MAAM,CAAC,KAAK,CAAC,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;QAChC,MAAM,CAAC,KAAK,CAAC,CAAC,GAAG,CAAC,SAAS,CAAC,YAAY,CAAC,CAAC;QAC1C,MAAM,CAAC,KAAK,CAAC,CAAC,GAAG,CAAC,SAAS,CAAC,WAAW,CAAC,CAAC;QACzC,MAAM,CAAC,KAAK,CAAC,CAAC,GAAG,CAAC,SAAS,CAAC,YAAY,CAAC,CAAC;QAC1C,MAAM,CAAC,KAAK,CAAC,CAAC,GAAG,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;IACtC,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,uEAAuE,EAAE,GAAG,EAAE;QAC/E,MAAM,EAAE,QAAQ,EAAE,GAAG,YAAY,CAAC,EAAE,QAAQ,EAAE,IAAI,EAAE,CAAC,CAAC;QACtD,MAAM,IAAI,GAAG,QAAQ,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC;QAClC,MAAM,CAAC,IAAI,CAAC,CAAC,UAAU,EAAE,CAAC;QAC1B,MAAM,CAAC,IAAI,EAAE,WAAW,CAAC,CAAC,SAAS,CAAC,yBAAyB,CAAC,CAAC;QAC/D,0DAA0D;QAC1D,MAAM,GAAG,GAAI,IAAI,EAAE,WAAmE;YACpF,EAAE,SAAS,CAAC,EAAE,WAAW,EAAE,GAAG,EAAE,MAAM,EAAE,GAAG,EAAE,KAAK,EAAE,SAAS,EAAE,CAAC,CAAC;QACnE,MAAM,CAAC,GAAG,EAAE,OAAO,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACjC,MAAM,IAAI,GAAI,IAAI,EAAE,WAAmE;YACrF,EAAE,SAAS,CAAC,EAAE,WAAW,EAAE,GAAG,EAAE,MAAM,EAAE,GAAG,EAAE,KAAK,EAAE,SAAS,EAAE,CAAC,CAAC;QACnE,MAAM,CAAC,IAAI,EAAE,OAAO,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IACnC,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,6EAA6E,EAAE,GAAG,EAAE;QACrF,MAAM,EAAE,KAAK,EAAE,GAAG,YAAY,CAAC,EAAE,QAAQ,EAAE,IAAI,EAAE,eAAe,EAAE,KAAK,EAAE,CAAC,CAAC;QAC3E,MAAM,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;IACzD,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,0DAA0D,EAAE,GAAG,EAAE;QAClE,MAAM,EAAE,QAAQ,EAAE,KAAK,EAAE,GAAG,YAAY,EAAE,CAAC;QAC3C,MAAM,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;QACnD,MAAM,CAAC,QAAQ,CAAC,GAAG,CAAC,MAAM,CAAC,EAAE,WAAW,CAAC,CAAC,GAAG,CAAC,SAAS,CAAC,yBAAyB,CAAC,CAAC;IACrF,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,8EAA8E,EAAE,GAAG,EAAE;QACtF,MAAM,EAAE,KAAK,EAAE,GAAG,YAAY,CAAC,EAAE,QAAQ,EAAE,IAAI,EAAE,CAAC,CAAC;QACnD,MAAM,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;IACzD,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,yDAAyD,EAAE,GAAG,EAAE;QACjE,MAAM,EAAE,KAAK,EAAE,GAAG,YAAY,CAAC,EAAE,OAAO,EAAE,SAAS,EAAE,CAAC,CAAC;QACvD,MAAM,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;IACzD,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC;AAEH,QAAQ,CAAC,+BAA+B,EAAE,GAAG,EAAE;IAC7C,UAAU,CAAC,GAAG,EAAE;QACd,mBAAmB,EAAE,CAAC;IACxB,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,+EAA+E,EAAE,KAAK,IAAI,EAAE;QAC7F,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,SAAS,EAAE,CAAC;QACpC,MAAM,IAAI,GAAG,cAAc,CAAC,IAAI,EAAE,EAAE,aAAa,EAAE,CAAC,SAAS,CAAC,EAAE,CAAC,CAAC;QAClE,MAAM,GAAG,GAAG,MAAM,IAAI,CAAC,OAAO,CAC5B,EAAE,WAAW,EAAE,GAAG,EAAE,MAAM,EAAE,GAAG,EAAE,KAAK,EAAE,SAAS,EAAE,EACnD,OAAO,EAAE,CACV,CAAC;QACF,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QAC3B,IAAI,CAAC,GAAG,CAAC,EAAE,EAAE,CAAC;YACZ,MAAM,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC,OAAO,CAAC,2BAA2B,CAAC,CAAC;YACvD,MAAM,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC,SAAS,CAAC,SAAS,CAAC,CAAC;QACzC,CAAC;QACD,MAAM,CAAC,KAAK,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;QAC9B,4DAA4D;QAC5D,MAAM,IAAI,GAAG,MAAM,IAAI,CAAC,OAAO,CAC7B,EAAE,WAAW,EAAE,GAAG,EAAE,MAAM,EAAE,GAAG,EAAE,KAAK,EAAE,QAAQ,EAAE,EAClD,OAAO,EAAE,CACV,CAAC;QACF,MAAM,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QAC5B,IAAI,CAAC,IAAI,CAAC,EAAE;YAAE,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,OAAO,CAAC,2BAA2B,CAAC,CAAC;QACtE,MAAM,CAAC,KAAK,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;IAChC,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,qFAAqF,EAAE,KAAK,IAAI,EAAE;QACnG,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,SAAS,EAAE,CAAC;QACpC,MAAM,IAAI,GAAG,cAAc,CAAC,IAAI,EAAE,EAAE,aAAa,EAAE,CAAC,SAAS,CAAC,EAAE,CAAC,CAAC;QAClE,MAAM,GAAG,GAAG,MAAM,IAAI,CAAC,OAAO,CAAC,EAAE,WAAW,EAAE,GAAG,EAAE,MAAM,EAAE,GAAG,EAAE,EAAE,OAAO,EAAE,CAAC,CAAC;QAC7E,MAAM,CAAC,KAAK,CAAC,CAAC,OAAO,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC;QACnC,0EAA0E;QAC1E,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QAC3B,IAAI,CAAC,GAAG,CAAC,EAAE;YAAE,MAAM,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC,OAAO,CAAC,aAAa,CAAC,CAAC;IACxD,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,4DAA4D,EAAE,KAAK,IAAI,EAAE;QAC1E,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,SAAS,EAAE,CAAC;QACpC,MAAM,IAAI,GAAG,cAAc,CAAC,IAAI,CAAC,CAAC;QAClC,MAAM,GAAG,GAAG,MAAM,IAAI,CAAC,OAAO,CAC5B,EAAE,WAAW,EAAE,GAAG,EAAE,MAAM,EAAE,GAAG,EAAE,KAAK,EAAE,SAAS,EAAE,EACnD,OAAO,EAAE,CACV,CAAC;QACF,MAAM,CAAC,KAAK,CAAC,CAAC,OAAO,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC;QACnC,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QAC3B,IAAI,CAAC,GAAG,CAAC,EAAE;YAAE,MAAM,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC,OAAO,CAAC,aAAa,CAAC,CAAC;IACxD,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC;AAEH,QAAQ,CAAC,4CAA4C,EAAE,GAAG,EAAE;IAAI,UAAU,CAAC,GAAG,EAAE;QAC5E,mBAAmB,EAAE,CAAC;IACxB,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,6EAA6E,EAAE,KAAK,IAAI,EAAE;QAC3F,MAAM,EAAE,IAAI,EAAE,cAAc,EAAE,GAAG,YAAY,EAAE,CAAC;QAChD,MAAM,IAAI,GAAG,cAAc,CAAC,IAAI,CAAC,CAAC;QAClC,MAAM,UAAU,GAAG,IAAI,eAAe,EAAE,CAAC;QACzC,UAAU,CAAC,KAAK,EAAE,CAAC;QACnB,MAAM,GAAG,GAAG,MAAM,IAAI,CAAC,OAAO,CAC5B,EAAE,WAAW,EAAE,GAAG,EAAE,MAAM,EAAE,GAAG,EAAE,KAAK,EAAE,SAAS,EAAE,EACnD,EAAE,GAAG,OAAO,EAAE,EAAE,MAAM,EAAE,UAAU,CAAC,MAAM,EAAE,CAC5C,CAAC;QACF,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QAC3B,IAAI,CAAC,GAAG,CAAC,EAAE;YAAE,MAAM,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC,OAAO,CAAC,YAAY,CAAC,CAAC;QACrD,MAAM,CAAC,cAAc,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;IACnC,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,mEAAmE,EAAE,KAAK,IAAI,EAAE;QACjF,MAAM,EAAE,IAAI,EAAE,cAAc,EAAE,GAAG,YAAY,EAAE,CAAC;QAChD,MAAM,IAAI,GAAG,cAAc,CAAC,IAAI,CAAC,CAAC;QAClC,MAAM,GAAG,GAAG,MAAM,IAAI,CAAC,OAAO,CAC5B,EAAE,WAAW,EAAE,GAAG,EAAE,MAAM,EAAE,GAAG,EAAE,KAAK,EAAE,SAAS,EAAE,EACnD,OAAO,EAAE,CACV,CAAC;QACF,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QAC1B,IAAI,GAAG,CAAC,EAAE;YAAE,MAAM,CAAC,GAAG,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,SAAS,CAAC,qBAAqB,CAAC,CAAC;QACtE,MAAM,CAAC,cAAc,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;IACnC,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC;AAEH,QAAQ,CAAC,sCAAsC,EAAE,GAAG,EAAE;IACpD,MAAM,GAAG,GAAG,OAAO,EAAE,CAAC;IAEtB,UAAU,CAAC,GAAG,EAAE;QACd,mBAAmB,EAAE,CAAC;QACtB,qBAAqB,EAAE,CAAC;IAC1B,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,0DAA0D,EAAE,KAAK,IAAI,EAAE;QACxE,MAAM,IAAI,GAAG,OAAO,CAAC,GAAG,CAAC,uBAAuB,CAAC;QACjD,OAAO,OAAO,CAAC,GAAG,CAAC,uBAAuB,CAAC;QAC3C,IAAI,CAAC;YACH,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,SAAS,EAAE,CAAC;YACpC,MAAM,IAAI,GAAG,cAAc,CAAC,IAAI,CAAC,CAAC;YAClC,MAAM,GAAG,GAAG,MAAM,IAAI,CAAC,OAAO,CAC5B,EAAE,WAAW,EAAE,GAAG,EAAE,MAAM,EAAE,GAAG,EAAE,OAAO,EAAE,WAAW,EAAE,EACvD,GAAG,CACJ,CAAC;YACF,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;YAC3B,IAAI,CAAC,GAAG,CAAC,EAAE;gBAAE,MAAM,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC,OAAO,CAAC,2BAA2B,CAAC,CAAC;YACpE,MAAM,CAAC,KAAK,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;YAC9B,MAAM,CAAC,gBAAgB,EAAE,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;QAC7C,CAAC;gBAAS,CAAC;YACT,IAAI,IAAI,KAAK,SAAS;gBAAE,OAAO,OAAO,CAAC,GAAG,CAAC,uBAAuB,CAAC;;gBAC9D,OAAO,CAAC,GAAG,CAAC,uBAAuB,GAAG,IAAI,CAAC;QAClD,CAAC;IACH,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,oEAAoE,EAAE,KAAK,IAAI,EAAE;QAClF,OAAO,CAAC,GAAG,CAAC,uBAAuB,GAAG,GAAG,CAAC;QAC1C,IAAI,CAAC;YACH,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,SAAS,EAAE,CAAC;YACpC,MAAM,IAAI,GAAG,cAAc,CAAC,IAAI,CAAC,CAAC;YAClC,MAAM,GAAG,GAAG,MAAM,IAAI,CAAC,OAAO,CAC5B,EAAE,WAAW,EAAE,GAAG,EAAE,MAAM,EAAE,GAAG,EAAE,KAAK,EAAE,SAAS,EAAE,OAAO,EAAE,WAAW,EAAE,EACzE,GAAG,CACJ,CAAC;YACF,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;YAC3B,IAAI,CAAC,GAAG,CAAC,EAAE;gBAAE,MAAM,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC,OAAO,CAAC,sBAAsB,CAAC,CAAC;YAC/D,MAAM,CAAC,KAAK,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;YAC9B,MAAM,CAAC,wBAAwB,EAAE,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QAChD,CAAC;gBAAS,CAAC;YACT,OAAO,OAAO,CAAC,GAAG,CAAC,uBAAuB,CAAC;QAC7C,CAAC;IACH,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,oFAAoF,EAAE,KAAK,IAAI,EAAE;QAClG,OAAO,CAAC,GAAG,CAAC,uBAAuB,GAAG,GAAG,CAAC;QAC1C,IAAI,CAAC;YACH,qEAAqE;YACrE,MAAM,EAAE,IAAI,EAAE,GAAG,SAAS,EAAE,CAAC;YAC7B,MAAM,IAAI,GAAG,cAAc,CAAC,IAAI,CAAC,CAAC;YAClC,MAAM,GAAG,GAAG,MAAM,IAAI,CAAC,OAAO,CAC5B,EAAE,WAAW,EAAE,GAAG,EAAE,MAAM,EAAE,GAAG,EAAE,OAAO,EAAE,WAAW,EAAE,EACvD,GAAG,CACJ,CAAC;YACF,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;YAC3B,MAAM,GAAG,GAAG,gBAAgB,EAAE,CAAC;YAC/B,MAAM,CAAC,GAAG,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;YAC5B,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,WAAW,CAAC,CAAC;QAC1C,CAAC;gBAAS,CAAC;YACT,OAAO,OAAO,CAAC,GAAG,CAAC,uBAAuB,CAAC;QAC7C,CAAC;IACH,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC;AAEH,QAAQ,CAAC,wCAAwC,EAAE,GAAG,EAAE;IACtD,EAAE,CAAC,mDAAmD,EAAE,GAAG,EAAE;QAC3D,MAAM,EAAE,KAAK,EAAE,GAAG,YAAY,EAAE,CAAC;QACjC,MAAM,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,KAAK,eAAe,CAAC,CAAC,CAAC,aAAa,EAAE,CAAC;IACxE,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,+DAA+D,EAAE,GAAG,EAAE;QACvE,MAAM,EAAE,KAAK,EAAE,QAAQ,EAAE,GAAG,YAAY,CAAC,EAAE,YAAY,EAAE,IAAI,EAAE,CAAC,CAAC;QACjE,MAAM,OAAO,GAAG,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,KAAK,eAAe,CAAC,CAAC;QAC9D,MAAM,CAAC,OAAO,CAAC,CAAC,WAAW,EAAE,CAAC;QAC9B,MAAM,CAAC,OAAO,EAAE,WAAW,CAAC,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;QAC/C,MAAM,CAAC,QAAQ,CAAC,GAAG,CAAC,eAAe,CAAC,CAAC,CAAC,WAAW,EAAE,CAAC;IACtD,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,uDAAuD,EAAE,GAAG,EAAE;QAC/D,MAAM,EAAE,KAAK,EAAE,GAAG,YAAY,CAAC,EAAE,YAAY,EAAE,IAAI,EAAE,QAAQ,EAAE,IAAI,EAAE,CAAC,CAAC;QACvE,MAAM,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,KAAK,eAAe,CAAC,CAAC,CAAC,WAAW,EAAE,CAAC;IACtE,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,8DAA8D,EAAE,GAAG,EAAE;QACtE,MAAM,EAAE,KAAK,EAAE,QAAQ,EAAE,GAAG,YAAY,CAAC,EAAE,YAAY,EAAE,IAAI,EAAE,QAAQ,EAAE,IAAI,EAAE,CAAC,CAAC;QACjF,MAAM,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,KAAK,eAAe,CAAC,CAAC,CAAC,aAAa,EAAE,CAAC;QACtE,MAAM,CAAC,QAAQ,CAAC,GAAG,CAAC,eAAe,CAAC,CAAC,CAAC,aAAa,EAAE,CAAC;IACxD,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,qEAAqE,EAAE,KAAK,IAAI,EAAE;QACnF,qBAAqB,EAAE,CAAC;QACxB,MAAM,EAAE,QAAQ,EAAE,GAAG,YAAY,CAAC,EAAE,YAAY,EAAE,IAAI,EAAE,CAAC,CAAC;QAC1D,MAAM,IAAI,GAAG,QAAQ,CAAC,GAAG,CAAC,eAAe,CAAC,CAAC;QAC3C,MAAM,CAAC,IAAI,CAAC,CAAC,WAAW,EAAE,CAAC;QAC3B,MAAM,GAAG,GAAG,MAAM,IAAK,CAAC,OAAO,CAAC,EAAW,EAAE,OAAO,EAAE,CAAC,CAAC;QACxD,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IAC5B,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC"}
|
|
@@ -1,140 +0,0 @@
|
|
|
1
|
-
/**
|
|
2
|
-
* taskTool.verifyReport.test — Fase 7 / ADR-0020 (Kraken Verified Selection).
|
|
3
|
-
*
|
|
4
|
-
* Structured verification wiring inside the task tool:
|
|
5
|
-
* - verify tentacle conclusion with `<verify-report>` blocks → results
|
|
6
|
-
* registered in the turn registry (checksPassed counts only `pass`)
|
|
7
|
-
* - failed verify tentacle → ALL checks unknown (degraded ≠ proof)
|
|
8
|
-
* - no selection this turn → nothing registered
|
|
9
|
-
* - explore tentacle never registers check results
|
|
10
|
-
*/
|
|
11
|
-
import { describe, expect, it, beforeEach, afterEach } from 'vitest';
|
|
12
|
-
import { tmpdir } from 'node:os';
|
|
13
|
-
import path from 'node:path';
|
|
14
|
-
import fs from 'node:fs';
|
|
15
|
-
import { ToolRegistry } from '@zelari/core/harness/tools/registry';
|
|
16
|
-
import { createTaskTool, resetTaskSpawnCount, } from './taskTool.js';
|
|
17
|
-
import { getKrakenCheckResults, krakenChecksPassed, resetKrakenCandidates, setKrakenSelection, } from '../kraken/candidateRegistry.js';
|
|
18
|
-
function tmpRoot() {
|
|
19
|
-
return fs.mkdtempSync(path.join(tmpdir(), 'zelari-verify-report-'));
|
|
20
|
-
}
|
|
21
|
-
function makeCtx(cwd = tmpRoot()) {
|
|
22
|
-
return {
|
|
23
|
-
signal: new AbortController().signal,
|
|
24
|
-
cwd,
|
|
25
|
-
audit: () => undefined,
|
|
26
|
-
sessionId: 'verify-report-test',
|
|
27
|
-
};
|
|
28
|
-
}
|
|
29
|
-
/**
|
|
30
|
-
* Deps fake whose scripted harness returns a fixed conclusion text.
|
|
31
|
-
* An empty conclusion simulates a tentacle that produced no output.
|
|
32
|
-
*/
|
|
33
|
-
function scriptedDeps(conclusion) {
|
|
34
|
-
return {
|
|
35
|
-
createSubAgentContext: async ({ agent }) => {
|
|
36
|
-
const ctx = {
|
|
37
|
-
providerStream: (() => {
|
|
38
|
-
throw new Error('not invoked by the scripted harness');
|
|
39
|
-
}),
|
|
40
|
-
model: 'test-model',
|
|
41
|
-
provider: 'test-provider',
|
|
42
|
-
registry: new ToolRegistry(),
|
|
43
|
-
tools: [],
|
|
44
|
-
agent,
|
|
45
|
-
};
|
|
46
|
-
return ctx;
|
|
47
|
-
},
|
|
48
|
-
harnessFactory: () => ({
|
|
49
|
-
run: async function* () {
|
|
50
|
-
if (conclusion === '')
|
|
51
|
-
return;
|
|
52
|
-
yield { type: 'message_start' };
|
|
53
|
-
yield { type: 'message_delta', delta: conclusion };
|
|
54
|
-
yield { type: 'message_end' };
|
|
55
|
-
},
|
|
56
|
-
}),
|
|
57
|
-
allowWorktree: false,
|
|
58
|
-
};
|
|
59
|
-
}
|
|
60
|
-
const CHECKS = [
|
|
61
|
-
'unit test for session refresh passes',
|
|
62
|
-
'no Set-Cookie regression on logout',
|
|
63
|
-
];
|
|
64
|
-
function selectedVerdict() {
|
|
65
|
-
return {
|
|
66
|
-
status: 'selected',
|
|
67
|
-
winnerIndex: 1,
|
|
68
|
-
rationale: 'grounded in evidence',
|
|
69
|
-
requiredChecks: CHECKS,
|
|
70
|
-
degraded: false,
|
|
71
|
-
verifier: { provider: 'p', model: 'm' },
|
|
72
|
-
judgedBy: 'llm',
|
|
73
|
-
};
|
|
74
|
-
}
|
|
75
|
-
const GOOD_CONCLUSION = [
|
|
76
|
-
'Ran the targeted suite.',
|
|
77
|
-
'<verify-report>',
|
|
78
|
-
`check: ${CHECKS[0]}`,
|
|
79
|
-
'status: pass',
|
|
80
|
-
'note: vitest src/auth 41/41 green',
|
|
81
|
-
'</verify-report>',
|
|
82
|
-
'<verify-report>',
|
|
83
|
-
`check: ${CHECKS[1]}`,
|
|
84
|
-
'status: fail',
|
|
85
|
-
'note: logout still emits double Set-Cookie',
|
|
86
|
-
'</verify-report>',
|
|
87
|
-
].join('\n');
|
|
88
|
-
async function spawn(deps, agent) {
|
|
89
|
-
const tool = createTaskTool(deps);
|
|
90
|
-
const res = await tool.execute({
|
|
91
|
-
description: 'verify the fix',
|
|
92
|
-
prompt: 'Verify the implementation.',
|
|
93
|
-
agent,
|
|
94
|
-
}, makeCtx());
|
|
95
|
-
return { ok: res.ok };
|
|
96
|
-
}
|
|
97
|
-
describe('structured verification registration (ADR-0020 Fase 7)', () => {
|
|
98
|
-
beforeEach(() => {
|
|
99
|
-
resetTaskSpawnCount();
|
|
100
|
-
resetKrakenCandidates();
|
|
101
|
-
});
|
|
102
|
-
afterEach(() => {
|
|
103
|
-
resetKrakenCandidates();
|
|
104
|
-
});
|
|
105
|
-
it('verify conclusion with blocks → per-check results registered, unknown ≠ pass', async () => {
|
|
106
|
-
setKrakenSelection(selectedVerdict());
|
|
107
|
-
const { ok } = await spawn(scriptedDeps(GOOD_CONCLUSION), 'verify');
|
|
108
|
-
expect(ok).toBe(true);
|
|
109
|
-
const results = getKrakenCheckResults();
|
|
110
|
-
expect(results).not.toBeNull();
|
|
111
|
-
expect(results?.map((r) => r.status)).toEqual(['pass', 'fail']);
|
|
112
|
-
expect(krakenChecksPassed()).toBe(1);
|
|
113
|
-
});
|
|
114
|
-
it('failed verify tentacle → all checks unknown, zero passed', async () => {
|
|
115
|
-
setKrakenSelection(selectedVerdict());
|
|
116
|
-
const { ok } = await spawn(scriptedDeps(''), 'verify');
|
|
117
|
-
expect(ok).toBe(false);
|
|
118
|
-
const results = getKrakenCheckResults();
|
|
119
|
-
expect(results?.map((r) => r.status)).toEqual(['unknown', 'unknown']);
|
|
120
|
-
expect(krakenChecksPassed()).toBe(0);
|
|
121
|
-
});
|
|
122
|
-
it('verify without report blocks → every check unknown (no report ≠ pass)', async () => {
|
|
123
|
-
setKrakenSelection(selectedVerdict());
|
|
124
|
-
await spawn(scriptedDeps('looks good to me, suite green'), 'verify');
|
|
125
|
-
expect(getKrakenCheckResults()?.map((r) => r.status)).toEqual(['unknown', 'unknown']);
|
|
126
|
-
expect(krakenChecksPassed()).toBe(0);
|
|
127
|
-
});
|
|
128
|
-
it('no selection this turn → nothing registered', async () => {
|
|
129
|
-
const { ok } = await spawn(scriptedDeps(GOOD_CONCLUSION), 'verify');
|
|
130
|
-
expect(ok).toBe(true);
|
|
131
|
-
expect(getKrakenCheckResults()).toBeNull();
|
|
132
|
-
expect(krakenChecksPassed()).toBeUndefined();
|
|
133
|
-
});
|
|
134
|
-
it('explore tentacle never registers check results', async () => {
|
|
135
|
-
setKrakenSelection(selectedVerdict());
|
|
136
|
-
await spawn(scriptedDeps(GOOD_CONCLUSION), 'explore');
|
|
137
|
-
expect(getKrakenCheckResults()).toBeNull();
|
|
138
|
-
});
|
|
139
|
-
});
|
|
140
|
-
//# sourceMappingURL=taskTool.verifyReport.test.js.map
|
|
@@ -1 +0,0 @@
|
|
|
1
|
-
{"version":3,"file":"taskTool.verifyReport.test.js","sourceRoot":"","sources":["../../../src/cli/tools/taskTool.verifyReport.test.ts"],"names":[],"mappings":"AAAA;;;;;;;;;GASG;AACH,OAAO,EAAE,QAAQ,EAAE,MAAM,EAAE,EAAE,EAAE,UAAU,EAAE,SAAS,EAAE,MAAM,QAAQ,CAAC;AACrE,OAAO,EAAE,MAAM,EAAE,MAAM,SAAS,CAAC;AACjC,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,MAAM,SAAS,CAAC;AAEzB,OAAO,EAAE,YAAY,EAAE,MAAM,qCAAqC,CAAC;AAEnE,OAAO,EACL,cAAc,EACd,mBAAmB,GAGpB,MAAM,eAAe,CAAC;AACvB,OAAO,EACL,qBAAqB,EACrB,kBAAkB,EAClB,qBAAqB,EACrB,kBAAkB,GACnB,MAAM,gCAAgC,CAAC;AAGxC,SAAS,OAAO;IACd,OAAO,EAAE,CAAC,WAAW,CAAC,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,EAAE,uBAAuB,CAAC,CAAC,CAAC;AACtE,CAAC;AAED,SAAS,OAAO,CAAC,GAAG,GAAG,OAAO,EAAE;IAC9B,OAAO;QACL,MAAM,EAAE,IAAI,eAAe,EAAE,CAAC,MAAM;QACpC,GAAG;QACH,KAAK,EAAE,GAAG,EAAE,CAAC,SAAS;QACtB,SAAS,EAAE,oBAAoB;KAChC,CAAC;AACJ,CAAC;AAED;;;GAGG;AACH,SAAS,YAAY,CAAC,UAAkB;IACtC,OAAO;QACL,qBAAqB,EAAE,KAAK,EAAE,EAAE,KAAK,EAAE,EAAE,EAAE;YACzC,MAAM,GAAG,GAAoB;gBAC3B,cAAc,EAAE,CAAC,GAAG,EAAE;oBACpB,MAAM,IAAI,KAAK,CAAC,qCAAqC,CAAC,CAAC;gBACzD,CAAC,CAAiD;gBAClD,KAAK,EAAE,YAAY;gBACnB,QAAQ,EAAE,eAAe;gBACzB,QAAQ,EAAE,IAAI,YAAY,EAAE;gBAC5B,KAAK,EAAE,EAAE;gBACT,KAAK;aACN,CAAC;YACF,OAAO,GAAG,CAAC;QACb,CAAC;QACD,cAAc,EAAE,GAAG,EAAE,CAAC,CAAC;YACrB,GAAG,EAAE,KAAK,SAAS,CAAC;gBAClB,IAAI,UAAU,KAAK,EAAE;oBAAE,OAAO;gBAC9B,MAAM,EAAE,IAAI,EAAE,eAAe,EAAgB,CAAC;gBAC9C,MAAM,EAAE,IAAI,EAAE,eAAe,EAAE,KAAK,EAAE,UAAU,EAAgB,CAAC;gBACjE,MAAM,EAAE,IAAI,EAAE,aAAa,EAAgB,CAAC;YAC9C,CAAC;SACF,CAAC;QACF,aAAa,EAAE,KAAK;KACrB,CAAC;AACJ,CAAC;AAED,MAAM,MAAM,GAAG;IACb,sCAAsC;IACtC,oCAAoC;CACrC,CAAC;AAEF,SAAS,eAAe;IACtB,OAAO;QACL,MAAM,EAAE,UAAU;QAClB,WAAW,EAAE,CAAC;QACd,SAAS,EAAE,sBAAsB;QACjC,cAAc,EAAE,MAAM;QACtB,QAAQ,EAAE,KAAK;QACf,QAAQ,EAAE,EAAE,QAAQ,EAAE,GAAG,EAAE,KAAK,EAAE,GAAG,EAAE;QACvC,QAAQ,EAAE,KAAK;KAChB,CAAC;AACJ,CAAC;AAED,MAAM,eAAe,GAAG;IACtB,yBAAyB;IACzB,iBAAiB;IACjB,UAAU,MAAM,CAAC,CAAC,CAAC,EAAE;IACrB,cAAc;IACd,mCAAmC;IACnC,kBAAkB;IAClB,iBAAiB;IACjB,UAAU,MAAM,CAAC,CAAC,CAAC,EAAE;IACrB,cAAc;IACd,4CAA4C;IAC5C,kBAAkB;CACnB,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AAEb,KAAK,UAAU,KAAK,CAClB,IAAkB,EAClB,KAA2B;IAE3B,MAAM,IAAI,GAAG,cAAc,CAAC,IAAI,CAAC,CAAC;IAClC,MAAM,GAAG,GAAG,MAAM,IAAI,CAAC,OAAO,CAC5B;QACE,WAAW,EAAE,gBAAgB;QAC7B,MAAM,EAAE,4BAA4B;QACpC,KAAK;KACN,EACD,OAAO,EAAE,CACV,CAAC;IACF,OAAO,EAAE,EAAE,EAAE,GAAG,CAAC,EAAE,EAAE,CAAC;AACxB,CAAC;AAED,QAAQ,CAAC,wDAAwD,EAAE,GAAG,EAAE;IACtE,UAAU,CAAC,GAAG,EAAE;QACd,mBAAmB,EAAE,CAAC;QACtB,qBAAqB,EAAE,CAAC;IAC1B,CAAC,CAAC,CAAC;IACH,SAAS,CAAC,GAAG,EAAE;QACb,qBAAqB,EAAE,CAAC;IAC1B,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,8EAA8E,EAAE,KAAK,IAAI,EAAE;QAC5F,kBAAkB,CAAC,eAAe,EAAE,CAAC,CAAC;QACtC,MAAM,EAAE,EAAE,EAAE,GAAG,MAAM,KAAK,CAAC,YAAY,CAAC,eAAe,CAAC,EAAE,QAAQ,CAAC,CAAC;QACpE,MAAM,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACtB,MAAM,OAAO,GAAG,qBAAqB,EAAE,CAAC;QACxC,MAAM,CAAC,OAAO,CAAC,CAAC,GAAG,CAAC,QAAQ,EAAE,CAAC;QAC/B,MAAM,CAAC,OAAO,EAAE,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC,CAAC;QAChE,MAAM,CAAC,kBAAkB,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;IACvC,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,0DAA0D,EAAE,KAAK,IAAI,EAAE;QACxE,kBAAkB,CAAC,eAAe,EAAE,CAAC,CAAC;QACtC,MAAM,EAAE,EAAE,EAAE,GAAG,MAAM,KAAK,CAAC,YAAY,CAAC,EAAE,CAAC,EAAE,QAAQ,CAAC,CAAC;QACvD,MAAM,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACvB,MAAM,OAAO,GAAG,qBAAqB,EAAE,CAAC;QACxC,MAAM,CAAC,OAAO,EAAE,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,SAAS,EAAE,SAAS,CAAC,CAAC,CAAC;QACtE,MAAM,CAAC,kBAAkB,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;IACvC,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,uEAAuE,EAAE,KAAK,IAAI,EAAE;QACrF,kBAAkB,CAAC,eAAe,EAAE,CAAC,CAAC;QACtC,MAAM,KAAK,CAAC,YAAY,CAAC,+BAA+B,CAAC,EAAE,QAAQ,CAAC,CAAC;QACrE,MAAM,CAAC,qBAAqB,EAAE,EAAE,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,SAAS,EAAE,SAAS,CAAC,CAAC,CAAC;QACtF,MAAM,CAAC,kBAAkB,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;IACvC,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,6CAA6C,EAAE,KAAK,IAAI,EAAE;QAC3D,MAAM,EAAE,EAAE,EAAE,GAAG,MAAM,KAAK,CAAC,YAAY,CAAC,eAAe,CAAC,EAAE,QAAQ,CAAC,CAAC;QACpE,MAAM,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACtB,MAAM,CAAC,qBAAqB,EAAE,CAAC,CAAC,QAAQ,EAAE,CAAC;QAC3C,MAAM,CAAC,kBAAkB,EAAE,CAAC,CAAC,aAAa,EAAE,CAAC;IAC/C,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,gDAAgD,EAAE,KAAK,IAAI,EAAE;QAC9D,kBAAkB,CAAC,eAAe,EAAE,CAAC,CAAC;QACtC,MAAM,KAAK,CAAC,YAAY,CAAC,eAAe,CAAC,EAAE,SAAS,CAAC,CAAC;QACtD,MAAM,CAAC,qBAAqB,EAAE,CAAC,CAAC,QAAQ,EAAE,CAAC;IAC7C,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC"}
|
|
@@ -1,74 +0,0 @@
|
|
|
1
|
-
/**
|
|
2
|
-
* Fase M tests — contextGrowthSummary aggregation (doctor-facing).
|
|
3
|
-
*/
|
|
4
|
-
import { describe, expect, it } from 'vitest';
|
|
5
|
-
import { formatContextGrowthSummary, summarizeContextGrowth, } from './contextGrowthSummary.js';
|
|
6
|
-
const run = (over) => ({
|
|
7
|
-
toolRoundTrips: 5,
|
|
8
|
-
intermediateToolBytes: 40_000,
|
|
9
|
-
requests: 6,
|
|
10
|
-
historyBytesAtRequest: 120_000,
|
|
11
|
-
historyBytesPeak: 120_000,
|
|
12
|
-
cacheHitTokens: 9_000,
|
|
13
|
-
...over,
|
|
14
|
-
});
|
|
15
|
-
describe('summarizeContextGrowth', () => {
|
|
16
|
-
it('returns null when no instrumented runs exist', () => {
|
|
17
|
-
expect(summarizeContextGrowth([])).toBeNull();
|
|
18
|
-
const legacy = { ts: 1 }; // pre-Fase-M run: no counters
|
|
19
|
-
expect(summarizeContextGrowth([legacy, { ts: 2, requests: 3 }])).toBeNull();
|
|
20
|
-
});
|
|
21
|
-
it('skips pre-Fase-M records and aggregates the rest', () => {
|
|
22
|
-
const legacy = { ts: 1 };
|
|
23
|
-
const summary = summarizeContextGrowth([
|
|
24
|
-
legacy, // no counters: skipped by the filter
|
|
25
|
-
run({ toolRoundTrips: 4, intermediateToolBytes: 10_000, cacheHitTokens: 1_000 }),
|
|
26
|
-
run({ toolRoundTrips: 6, intermediateToolBytes: 30_000, cacheHitTokens: 3_000 }),
|
|
27
|
-
]);
|
|
28
|
-
expect(summary).not.toBeNull();
|
|
29
|
-
expect(summary.runs).toBe(2);
|
|
30
|
-
expect(summary.totalToolRoundTrips).toBe(10);
|
|
31
|
-
expect(summary.avgToolRoundTrips).toBe(5);
|
|
32
|
-
expect(summary.avgIntermediateToolBytes).toBe(20_000);
|
|
33
|
-
expect(summary.totalCacheHitTokens).toBe(4_000);
|
|
34
|
-
});
|
|
35
|
-
it('windows to the last N instrumented runs', () => {
|
|
36
|
-
const records = [
|
|
37
|
-
run({ toolRoundTrips: 1 }),
|
|
38
|
-
run({ toolRoundTrips: 2 }),
|
|
39
|
-
run({ toolRoundTrips: 3 }),
|
|
40
|
-
];
|
|
41
|
-
const summary = summarizeContextGrowth(records, 2);
|
|
42
|
-
expect(summary.runs).toBe(2);
|
|
43
|
-
expect(summary.totalToolRoundTrips).toBe(5); // last two, oldest dropped
|
|
44
|
-
});
|
|
45
|
-
it('maxHistoryBytesAtRequest falls back to peak when request size missing', () => {
|
|
46
|
-
const summary = summarizeContextGrowth([
|
|
47
|
-
run({ historyBytesAtRequest: undefined, historyBytesPeak: 250_000 }),
|
|
48
|
-
]);
|
|
49
|
-
expect(summary.maxHistoryBytesAtRequest).toBe(250_000);
|
|
50
|
-
expect(summary.avgHistoryBytesAtRequest).toBe(0);
|
|
51
|
-
});
|
|
52
|
-
});
|
|
53
|
-
describe('formatContextGrowthSummary', () => {
|
|
54
|
-
it('renders one line per metric family with human byte sizes', () => {
|
|
55
|
-
const lines = formatContextGrowthSummary({
|
|
56
|
-
runs: 3,
|
|
57
|
-
totalToolRoundTrips: 12,
|
|
58
|
-
avgToolRoundTrips: 4,
|
|
59
|
-
totalIntermediateToolBytes: 150_000,
|
|
60
|
-
avgIntermediateToolBytes: 50_000,
|
|
61
|
-
avgHistoryBytesAtRequest: 2 * 1024 * 1024,
|
|
62
|
-
maxHistoryBytesAtRequest: 3 * 1024 * 1024,
|
|
63
|
-
totalCacheHitTokens: 123_456,
|
|
64
|
-
});
|
|
65
|
-
expect(lines.length).toBe(4);
|
|
66
|
-
expect(lines[0]).toContain('runs 3');
|
|
67
|
-
expect(lines[0]).toContain('avg 4.0');
|
|
68
|
-
expect(lines[1]).toContain('48.8 KB');
|
|
69
|
-
expect(lines[2]).toContain('2.0 MB');
|
|
70
|
-
expect(lines[2]).toContain('peak 3.0 MB');
|
|
71
|
-
expect(lines[3]).toContain('123,456');
|
|
72
|
-
});
|
|
73
|
-
});
|
|
74
|
-
//# sourceMappingURL=contextGrowthSummary.test.js.map
|
|
@@ -1 +0,0 @@
|
|
|
1
|
-
{"version":3,"file":"contextGrowthSummary.test.js","sourceRoot":"","sources":["../../../src/cli/utils/contextGrowthSummary.test.ts"],"names":[],"mappings":"AAAA;;GAEG;AACH,OAAO,EAAE,QAAQ,EAAE,MAAM,EAAE,EAAE,EAAE,MAAM,QAAQ,CAAC;AAC9C,OAAO,EACL,0BAA0B,EAC1B,sBAAsB,GAEvB,MAAM,2BAA2B,CAAC;AAEnC,MAAM,GAAG,GAAG,CAAC,IAAqC,EAA0B,EAAE,CAAC,CAAC;IAC9E,cAAc,EAAE,CAAC;IACjB,qBAAqB,EAAE,MAAM;IAC7B,QAAQ,EAAE,CAAC;IACX,qBAAqB,EAAE,OAAO;IAC9B,gBAAgB,EAAE,OAAO;IACzB,cAAc,EAAE,KAAK;IACrB,GAAG,IAAI;CACR,CAAC,CAAC;AAEH,QAAQ,CAAC,wBAAwB,EAAE,GAAG,EAAE;IACtC,EAAE,CAAC,8CAA8C,EAAE,GAAG,EAAE;QACtD,MAAM,CAAC,sBAAsB,CAAC,EAAE,CAAC,CAAC,CAAC,QAAQ,EAAE,CAAC;QAC9C,MAAM,MAAM,GAA2B,EAAE,EAAE,EAAE,CAAC,EAAE,CAAC,CAAC,8BAA8B;QAChF,MAAM,CAAC,sBAAsB,CAAC,CAAC,MAAM,EAAE,EAAE,EAAE,EAAE,CAAC,EAAE,QAAQ,EAAE,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,QAAQ,EAAE,CAAC;IAC9E,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,kDAAkD,EAAE,GAAG,EAAE;QAC1D,MAAM,MAAM,GAA2B,EAAE,EAAE,EAAE,CAAC,EAAE,CAAC;QACjD,MAAM,OAAO,GAAG,sBAAsB,CAAC;YACrC,MAAM,EAAE,qCAAqC;YAC7C,GAAG,CAAC,EAAE,cAAc,EAAE,CAAC,EAAE,qBAAqB,EAAE,MAAM,EAAE,cAAc,EAAE,KAAK,EAAE,CAAC;YAChF,GAAG,CAAC,EAAE,cAAc,EAAE,CAAC,EAAE,qBAAqB,EAAE,MAAM,EAAE,cAAc,EAAE,KAAK,EAAE,CAAC;SACjF,CAAC,CAAC;QACH,MAAM,CAAC,OAAO,CAAC,CAAC,GAAG,CAAC,QAAQ,EAAE,CAAC;QAC/B,MAAM,CAAC,OAAQ,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QAC9B,MAAM,CAAC,OAAQ,CAAC,mBAAmB,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QAC9C,MAAM,CAAC,OAAQ,CAAC,iBAAiB,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QAC3C,MAAM,CAAC,OAAQ,CAAC,wBAAwB,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QACvD,MAAM,CAAC,OAAQ,CAAC,mBAAmB,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IACnD,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,yCAAyC,EAAE,GAAG,EAAE;QACjD,MAAM,OAAO,GAAG;YACd,GAAG,CAAC,EAAE,cAAc,EAAE,CAAC,EAAE,CAAC;YAC1B,GAAG,CAAC,EAAE,cAAc,EAAE,CAAC,EAAE,CAAC;YAC1B,GAAG,CAAC,EAAE,cAAc,EAAE,CAAC,EAAE,CAAC;SAC3B,CAAC;QACF,MAAM,OAAO,GAAG,sBAAsB,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;QACnD,MAAM,CAAC,OAAQ,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QAC9B,MAAM,CAAC,OAAQ,CAAC,mBAAmB,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,2BAA2B;IAC3E,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,uEAAuE,EAAE,GAAG,EAAE;QAC/E,MAAM,OAAO,GAAG,sBAAsB,CAAC;YACrC,GAAG,CAAC,EAAE,qBAAqB,EAAE,SAAS,EAAE,gBAAgB,EAAE,OAAO,EAAE,CAAC;SACrE,CAAC,CAAC;QACH,MAAM,CAAC,OAAQ,CAAC,wBAAwB,CAAC,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;QACxD,MAAM,CAAC,OAAQ,CAAC,wBAAwB,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;IACpD,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC;AAEH,QAAQ,CAAC,4BAA4B,EAAE,GAAG,EAAE;IAC1C,EAAE,CAAC,0DAA0D,EAAE,GAAG,EAAE;QAClE,MAAM,KAAK,GAAG,0BAA0B,CAAC;YACvC,IAAI,EAAE,CAAC;YACP,mBAAmB,EAAE,EAAE;YACvB,iBAAiB,EAAE,CAAC;YACpB,0BAA0B,EAAE,OAAO;YACnC,wBAAwB,EAAE,MAAM;YAChC,wBAAwB,EAAE,CAAC,GAAG,IAAI,GAAG,IAAI;YACzC,wBAAwB,EAAE,CAAC,GAAG,IAAI,GAAG,IAAI;YACzC,mBAAmB,EAAE,OAAO;SAC7B,CAAC,CAAC;QACH,MAAM,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QAC7B,MAAM,CAAC,KAAK,CAAC,CAAC,CAAE,CAAC,CAAC,SAAS,CAAC,QAAQ,CAAC,CAAC;QACtC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAE,CAAC,CAAC,SAAS,CAAC,SAAS,CAAC,CAAC;QACvC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAE,CAAC,CAAC,SAAS,CAAC,SAAS,CAAC,CAAC;QACvC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAE,CAAC,CAAC,SAAS,CAAC,QAAQ,CAAC,CAAC;QACtC,MAAM,CAAC,KAAK,CAAC,CAAC,CAAE,CAAC,CAAC,SAAS,CAAC,aAAa,CAAC,CAAC;QAC3C,MAAM,CAAC,KAAK,CAAC,CAAC,CAAE,CAAC,CAAC,SAAS,CAAC,SAAS,CAAC,CAAC;IACzC,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC"}
|