@r1z4x/tezgah 1.11.0 → 1.14.0
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/CHANGELOG.md +507 -1
- package/MANIFEST +40 -0
- package/NOTICE +152 -0
- package/README.de.md +41 -20
- package/README.es.md +39 -19
- package/README.fr.md +41 -20
- package/README.md +18 -10
- package/README.tr.md +39 -21
- package/bin/codegen +9 -2
- package/bin/consult +6 -0
- package/bin/tezgah-adhd +3 -0
- package/bin/tezgah-agents +3 -0
- package/bin/tezgah-capture +3 -0
- package/bin/tezgah-design +329 -54
- package/bin/tezgah-docs +158 -42
- package/bin/tezgah-dsh +1 -1
- package/bin/tezgah-gate +7 -2
- package/bin/tezgah-import-ai-research +16 -4
- package/bin/tezgah-index +3 -0
- package/bin/tezgah-lessons +13 -4
- package/bin/tezgah-mcp +4 -0
- package/bin/tezgah-migrate +3 -0
- package/bin/tezgah-ocr-review +2 -1
- package/bin/tezgah-optimize +4 -1
- package/bin/tezgah-pony +3 -0
- package/bin/tezgah-research +51 -10
- package/bin/tezgah-rollback +3 -0
- package/bin/tezgah-route +1 -1
- package/bin/tezgah-setup +1741 -95
- package/bin/tezgah-status +11 -1
- package/bin/tezgah-task +58 -17
- package/bin/tezgah.js +10 -3
- package/commands/adhd.md +1 -1
- package/docs/README.md +19 -8
- package/docs/architecture.md +2 -2
- package/docs/citations-baseline.json +11 -17
- package/docs/clarity-baseline.json +12 -6
- package/docs/contract.md +83 -34
- package/docs/environment.md +59 -24
- package/docs/evidence.md +128 -25
- package/docs/gate.md +77 -23
- package/docs/glossary.md +1 -1
- package/docs/hosts.md +343 -48
- package/docs/layers.md +2 -2
- package/docs/models.md +6 -0
- package/docs/operations.md +83 -29
- package/docs/research.md +37 -11
- package/docs/skills.md +31 -8
- package/docs/status-line.md +39 -23
- package/docs/testing.md +50 -17
- package/docs/vision.md +6 -2
- package/hooks/hooks.json +1 -1
- package/hooks/lang/en.json +16 -2
- package/hooks/lang/tr.json +14 -2
- package/hooks/projects-posttooluse.py +11 -3
- package/hooks/projects-stop.py +3 -2
- package/hooks/tezgah_agents.py +27 -3
- package/hooks/tezgah_attest.py +79 -10
- package/hooks/tezgah_context.py +61 -31
- package/hooks/tezgah_gate.py +480 -126
- package/hooks/tezgah_integrity.py +577 -133
- package/hooks/tezgah_lessons.py +5 -2
- package/hooks/tezgah_paths.py +60 -8
- package/hooks/tezgah_policy.py +48 -18
- package/hooks/tezgah_replay.py +8 -2
- package/hooks/tezgah_research.py +167 -24
- package/hooks/tezgah_skill_pick.py +1 -1
- package/hooks/tezgah_snapshot.py +6 -3
- package/hooks/tezgah_stack.py +236 -1
- package/hooks/tezgah_stopspec.py +2 -1
- package/hooks/tezgah_task.py +2 -1
- package/hosts/amp/hook.py +217 -0
- package/hosts/amp/tezgah-plugin.ts.in +396 -0
- package/hosts/antigravity/hook.py +255 -0
- package/hosts/codex/hook.py +3 -2
- package/hosts/commandcode/hook.py +239 -0
- package/hosts/copilot/hook.py +284 -0
- package/hosts/cursor/hook.py +4 -2
- package/hosts/droid/hook.py +259 -0
- package/hosts/dsh/hooks.json +1 -1
- package/hosts/gemini/hook.py +213 -0
- package/hosts/goose/hook.py +145 -0
- package/hosts/omp/hook.py +5 -2
- package/hosts/omp/tezgah-hook.ts.in +157 -42
- package/hosts/opencode/plugins/tezgah.js +209 -63
- package/hosts/qwen/hook.py +206 -0
- package/output-styles/tezgah.md +11 -6
- package/package.json +6 -3
- package/packaging/build.sh +4 -4
- package/packaging/install.sh +11 -0
- package/packaging/upgrade.sh +3 -0
- package/skills/ai-research/LICENSE +21 -0
- package/skills/design-contract/SKILL.md +27 -6
- package/skills/design-library/LICENSE +21 -0
- package/skills/harness/SKILL.md +2 -0
- package/skills/i-have-adhd/SKILL.md +10 -1
- package/skills/plan-add/SKILL.md +5 -1
- package/skills/plan-add/render_table.py +7 -0
- package/skills/pm-frameworks/LICENSE +21 -0
- package/skills/research/SKILL.md +22 -9
- package/skills/rl-env/LICENSE +201 -0
- package/skills/tezgah-contract/SKILL.md +21 -8
- package/statusline.py +20 -9
- package/third_party/open-code-review/LICENSE +201 -0
- package/third_party/open-code-review/README.md +15 -0
- package/third_party/open-code-review/prompts/main_task_system.md +26 -0
- package/third_party/open-code-review/prompts/review_filter_task_system.md +13 -0
- package/workflows/graph-map.js +14 -5
- package/workflows/graph-review.js +14 -2
package/MANIFEST
CHANGED
|
@@ -161,19 +161,28 @@ hooks/tezgah_taste.py
|
|
|
161
161
|
hooks/tezgah_taste_ledger.py
|
|
162
162
|
hooks/tezgah_untrusted.py
|
|
163
163
|
hooks/tezgah_update.py
|
|
164
|
+
hosts/amp/hook.py
|
|
165
|
+
hosts/amp/tezgah-plugin.ts.in
|
|
166
|
+
hosts/antigravity/hook.py
|
|
164
167
|
hosts/codex/hook.py
|
|
165
168
|
hosts/codex/hooks.json
|
|
169
|
+
hosts/commandcode/hook.py
|
|
170
|
+
hosts/copilot/hook.py
|
|
166
171
|
hosts/cursor/hook.py
|
|
167
172
|
hosts/cursor/hooks.json
|
|
173
|
+
hosts/droid/hook.py
|
|
168
174
|
hosts/dsh/hooks.json
|
|
169
175
|
hosts/dsh/statusline/lib/client.js
|
|
170
176
|
hosts/dsh/statusline/lib/index.js
|
|
171
177
|
hosts/dsh/statusline/package.json
|
|
178
|
+
hosts/gemini/hook.py
|
|
179
|
+
hosts/goose/hook.py
|
|
172
180
|
hosts/omp/hook.py
|
|
173
181
|
hosts/omp/tezgah-hook.ts.in
|
|
174
182
|
hosts/opencode/plugins/tezgah.js
|
|
175
183
|
hosts/opencode/tui/cli/tui.tsx
|
|
176
184
|
hosts/opencode/tui/tezgah-tui.tsx
|
|
185
|
+
hosts/qwen/hook.py
|
|
177
186
|
output-styles/tezgah.md
|
|
178
187
|
package.json
|
|
179
188
|
packaging/build.sh
|
|
@@ -547,6 +556,7 @@ skills/ai-research/22-agent-native-research-artifact/research-manager/references
|
|
|
547
556
|
skills/ai-research/22-agent-native-research-artifact/research-manager/references/session-protocol.md
|
|
548
557
|
skills/ai-research/22-agent-native-research-artifact/rigor-reviewer/ENTRY.md
|
|
549
558
|
skills/ai-research/22-agent-native-research-artifact/rigor-reviewer/references/review-dimensions.md
|
|
559
|
+
skills/ai-research/LICENSE
|
|
550
560
|
skills/ai-research/SKILL.md
|
|
551
561
|
skills/ai-research/SOURCE
|
|
552
562
|
skills/ai-research/index/1-frame.md
|
|
@@ -563,6 +573,7 @@ skills/design-contract/SKILL.md
|
|
|
563
573
|
skills/design-contract/data-views.md
|
|
564
574
|
skills/design-library/EVALS.md
|
|
565
575
|
skills/design-library/INDEX.md
|
|
576
|
+
skills/design-library/LICENSE
|
|
566
577
|
skills/design-library/SKILL.md
|
|
567
578
|
skills/design-library/SOURCE
|
|
568
579
|
skills/design-library/adaptive-interfaces/skills/colour-independence/ENTRY.md
|
|
@@ -799,6 +810,7 @@ skills/plan-status/EVALS.md
|
|
|
799
810
|
skills/plan-status/SKILL.md
|
|
800
811
|
skills/plan-sync/EVALS.md
|
|
801
812
|
skills/plan-sync/SKILL.md
|
|
813
|
+
skills/pm-frameworks/LICENSE
|
|
802
814
|
skills/pm-frameworks/SKILL.md
|
|
803
815
|
skills/pm-frameworks/SOURCE
|
|
804
816
|
skills/pm-frameworks/intended-vs-implemented/ENTRY.md
|
|
@@ -811,6 +823,7 @@ skills/product-flow/EVALS.md
|
|
|
811
823
|
skills/product-flow/SKILL.md
|
|
812
824
|
skills/research/EVALS.md
|
|
813
825
|
skills/research/SKILL.md
|
|
826
|
+
skills/rl-env/LICENSE
|
|
814
827
|
skills/rl-env/SKILL.md
|
|
815
828
|
skills/rl-env/SOURCE
|
|
816
829
|
skills/rl-env/generate-nemo-gym-env/ENTRY.md
|
|
@@ -830,6 +843,7 @@ skills/rl-env/rl-env-from-description/references/verifiers.md
|
|
|
830
843
|
skills/tezgah-contract/EVALS.md
|
|
831
844
|
skills/tezgah-contract/SKILL.md
|
|
832
845
|
statusline.py
|
|
846
|
+
tests/_amp_plugin_harness.mjs
|
|
833
847
|
tests/_fuzz_shell_reader.mjs
|
|
834
848
|
tests/_mcp_stdio.py
|
|
835
849
|
tests/_omp_extension_harness.mjs
|
|
@@ -849,10 +863,12 @@ tests/e2e_analyze_mobile.py
|
|
|
849
863
|
tests/e2e_analyze_web.py
|
|
850
864
|
tests/e2e_analyze_wiring.py
|
|
851
865
|
tests/e2e_docker_cycle.py
|
|
866
|
+
tests/e2e_droid.py
|
|
852
867
|
tests/e2e_dsh_statusline.py
|
|
853
868
|
tests/e2e_omp_statusline.py
|
|
854
869
|
tests/e2e_packaged_install.py
|
|
855
870
|
tests/e2e_plan_flow.py
|
|
871
|
+
tests/e2e_qwen.py
|
|
856
872
|
tests/e2e_tezgah_mcp.py
|
|
857
873
|
tests/fixtures/rules8-known-truth/bodies/arxiv-1.xml
|
|
858
874
|
tests/fixtures/rules8-known-truth/bodies/arxiv-10.xml
|
|
@@ -906,6 +922,9 @@ tests/stop_spec_eval.py
|
|
|
906
922
|
tests/support.py
|
|
907
923
|
tests/test_agents.py
|
|
908
924
|
tests/test_ai_research_library.py
|
|
925
|
+
tests/test_amp_hook.py
|
|
926
|
+
tests/test_amp_plugin.py
|
|
927
|
+
tests/test_antigravity_hook.py
|
|
909
928
|
tests/test_apps_registry.py
|
|
910
929
|
tests/test_attest.py
|
|
911
930
|
tests/test_bash_agreement.py
|
|
@@ -914,32 +933,46 @@ tests/test_cheap_source.py
|
|
|
914
933
|
tests/test_clarity.py
|
|
915
934
|
tests/test_claude_adapters.py
|
|
916
935
|
tests/test_claude_registration.py
|
|
936
|
+
tests/test_cli_flags.py
|
|
917
937
|
tests/test_codegen.py
|
|
918
938
|
tests/test_codex_hook.py
|
|
939
|
+
tests/test_commandcode_hook.py
|
|
919
940
|
tests/test_components.py
|
|
920
941
|
tests/test_constraint_pinning.py
|
|
921
942
|
tests/test_consult_arena.py
|
|
922
943
|
tests/test_context.py
|
|
923
944
|
tests/test_context_diet.py
|
|
945
|
+
tests/test_context_switch_text.py
|
|
924
946
|
tests/test_control_plane.py
|
|
947
|
+
tests/test_copilot_hook.py
|
|
925
948
|
tests/test_cursor_cap.py
|
|
926
949
|
tests/test_cursor_hook.py
|
|
927
950
|
tests/test_decisions.py
|
|
928
951
|
tests/test_design_cli.py
|
|
929
952
|
tests/test_docs.py
|
|
953
|
+
tests/test_docs_citations.py
|
|
954
|
+
tests/test_docs_reach.py
|
|
930
955
|
tests/test_docs_router.py
|
|
931
956
|
tests/test_doctor.py
|
|
957
|
+
tests/test_droid_hook.py
|
|
932
958
|
tests/test_dsh_hooks.py
|
|
933
959
|
tests/test_dsh_launcher.py
|
|
934
960
|
tests/test_dsh_statusline.py
|
|
961
|
+
tests/test_e2e_omp_statusline.py
|
|
935
962
|
tests/test_embed.py
|
|
936
963
|
tests/test_encoding.py
|
|
937
964
|
tests/test_enforcers.py
|
|
965
|
+
tests/test_env_vars_audit.py
|
|
966
|
+
tests/test_evidence_docs.py
|
|
938
967
|
tests/test_evidence_store.py
|
|
939
968
|
tests/test_extension.py
|
|
940
969
|
tests/test_fuzz_shell.py
|
|
941
970
|
tests/test_gate.py
|
|
971
|
+
tests/test_gate_rules_audit.py
|
|
972
|
+
tests/test_gate_shell_audit.py
|
|
942
973
|
tests/test_gate_vectors.py
|
|
974
|
+
tests/test_gemini_hook.py
|
|
975
|
+
tests/test_goose_hook.py
|
|
943
976
|
tests/test_graph_verbs.py
|
|
944
977
|
tests/test_graph_workflows.py
|
|
945
978
|
tests/test_guard.py
|
|
@@ -961,12 +994,14 @@ tests/test_lesson_lineage.py
|
|
|
961
994
|
tests/test_lessons.py
|
|
962
995
|
tests/test_lessons_compile.py
|
|
963
996
|
tests/test_lifecycle_sweep.py
|
|
997
|
+
tests/test_links.py
|
|
964
998
|
tests/test_live_probe.py
|
|
965
999
|
tests/test_mcp_features.py
|
|
966
1000
|
tests/test_mcp_server.py
|
|
967
1001
|
tests/test_migrate.py
|
|
968
1002
|
tests/test_models.py
|
|
969
1003
|
tests/test_neuter_matrix.py
|
|
1004
|
+
tests/test_npm_shim_probe.py
|
|
970
1005
|
tests/test_ocr_review.py
|
|
971
1006
|
tests/test_omp_extension.py
|
|
972
1007
|
tests/test_omp_hook.py
|
|
@@ -981,6 +1016,7 @@ tests/test_platform.py
|
|
|
981
1016
|
tests/test_product_flow_e2e.py
|
|
982
1017
|
tests/test_product_next.py
|
|
983
1018
|
tests/test_providers.py
|
|
1019
|
+
tests/test_qwen_hook.py
|
|
984
1020
|
tests/test_rank.py
|
|
985
1021
|
tests/test_replay.py
|
|
986
1022
|
tests/test_replay_context.py
|
|
@@ -994,6 +1030,7 @@ tests/test_research_seal.py
|
|
|
994
1030
|
tests/test_research_stats.py
|
|
995
1031
|
tests/test_review_quote.py
|
|
996
1032
|
tests/test_setup.py
|
|
1033
|
+
tests/test_setup_farm.py
|
|
997
1034
|
tests/test_shapes.py
|
|
998
1035
|
tests/test_skill_pick.py
|
|
999
1036
|
tests/test_skill_search.py
|
|
@@ -1004,6 +1041,8 @@ tests/test_stack.py
|
|
|
1004
1041
|
tests/test_statusline.py
|
|
1005
1042
|
tests/test_stop_after_block.py
|
|
1006
1043
|
tests/test_stopspec.py
|
|
1044
|
+
tests/test_store_concurrency.py
|
|
1045
|
+
tests/test_switch_latch_subagent.py
|
|
1007
1046
|
tests/test_switch_latching.py
|
|
1008
1047
|
tests/test_task.py
|
|
1009
1048
|
tests/test_taste.py
|
|
@@ -1011,6 +1050,7 @@ tests/test_taste_cli.py
|
|
|
1011
1050
|
tests/test_taste_ledger.py
|
|
1012
1051
|
tests/test_tezgah_mcp_wiring.py
|
|
1013
1052
|
tests/test_triage.py
|
|
1053
|
+
tests/test_uninstall_roundtrip.py
|
|
1014
1054
|
tests/test_untrusted.py
|
|
1015
1055
|
tests/test_update.py
|
|
1016
1056
|
third_party/open-code-review/LICENSE
|
package/NOTICE
ADDED
|
@@ -0,0 +1,152 @@
|
|
|
1
|
+
Third-party material vendored into or adapted by this repository. The MIT LICENSE
|
|
2
|
+
at the root covers tezgah's own files; the copies below keep their original terms.
|
|
3
|
+
Each vendored tree (a directory with a SOURCE file) carries upstream's LICENSE
|
|
4
|
+
beside its SOURCE, verbatim at the revision SOURCE names.
|
|
5
|
+
|
|
6
|
+
skills/no-ai-slop - MIT, Peter Yang, https://github.com/petergyang/no-ai-slop
|
|
7
|
+
skills/ponytail - MIT. Origin not recorded at vendoring time; if you are the
|
|
8
|
+
author, open an issue and the attribution will be corrected.
|
|
9
|
+
skills/i-have-adhd - MIT, ayghri, https://github.com/ayghri/i-have-adhd.
|
|
10
|
+
Adapted for tezgah: the ten rules are kept and two are
|
|
11
|
+
rewritten (the state restatement points at the todo list
|
|
12
|
+
instead of duplicating it, and a time estimate is marked
|
|
13
|
+
as an estimate), the user-invocation frontmatter is
|
|
14
|
+
dropped so the skill is model-invocable from the router,
|
|
15
|
+
and the trigger phrases are tezgah's.
|
|
16
|
+
skills/research - MIT, Orchestra Research. Adapts the orchestration layer of
|
|
17
|
+
its AI-research-SKILLs library (the autoresearch,
|
|
18
|
+
ara-rigor-reviewer and ara-research-manager skills):
|
|
19
|
+
https://github.com/orchestra-research/AI-research-SKILLs
|
|
20
|
+
Nothing is copied verbatim.
|
|
21
|
+
skills/product-analysis - MIT. Adapts the intended-vs-implemented method (a gap
|
|
22
|
+
is a finding only when the documented intent and the code
|
|
23
|
+
that enforces it are both cited) and the opportunity-score
|
|
24
|
+
framing from https://github.com/phuryn/pm-skills, whose
|
|
25
|
+
pm-ai-shipping and pm-product-discovery skills attribute
|
|
26
|
+
those in turn to Teresa Torres, Dan Olsen and Alberto
|
|
27
|
+
Savoia. Nothing is copied verbatim: the evidence classes,
|
|
28
|
+
the two-axis rule, the artifact shape and the scorecard are
|
|
29
|
+
tezgah's, and the skill points at the marketplace by URL
|
|
30
|
+
instead of reproducing its text.
|
|
31
|
+
|
|
32
|
+
skills/pm-frameworks - MIT, Pawel Huryn, https://github.com/phuryn/pm-skills.
|
|
33
|
+
VENDORED at revision
|
|
34
|
+
8607e3b077817f89bf4a9b623246219734ac3be0 (2026-09-19): two
|
|
35
|
+
skills, byte-for-byte upstream -
|
|
36
|
+
`intended-vs-implemented/ENTRY.md` and
|
|
37
|
+
`opportunity-solution-tree/ENTRY.md`, each upstream's
|
|
38
|
+
`SKILL.md` renamed so a host that lists every `SKILL.md`
|
|
39
|
+
below its skills directory does not list it. The entry point
|
|
40
|
+
`skills/pm-frameworks/SKILL.md` is tezgah's own and is not
|
|
41
|
+
part of the manifest; the other 67 skills and the plugin
|
|
42
|
+
manifest are not copied. `skills/pm-frameworks/SOURCE`
|
|
43
|
+
records the upstream paths and the per-file sha256, and
|
|
44
|
+
`tests/test_skills.py` holds the tree to it.
|
|
45
|
+
|
|
46
|
+
skills/ai-research - MIT, Orchestra Research. VENDORED from the same library at
|
|
47
|
+
revision 773a52944ba4747a18bd4ae9ade53fff041adcbc
|
|
48
|
+
(2026-06-15): 98 skills in 23 categories, each with its
|
|
49
|
+
markdown references, in the upstream layout. The bodies are
|
|
50
|
+
byte-for-byte upstream, frontmatter included - 96 keep
|
|
51
|
+
Orchestra Research's `author` line, two
|
|
52
|
+
(`ml-training-recipes`, `a-evolve`) keep the contributing
|
|
53
|
+
author's - and a markdown file with no frontmatter carries
|
|
54
|
+
one attribution comment. Each entry's `SKILL.md` sits at
|
|
55
|
+
the upstream path renamed `ENTRY.md`, bytes unchanged, so a
|
|
56
|
+
host that lists every `SKILL.md` below its skills directory
|
|
57
|
+
does not list it.
|
|
58
|
+
Three machine-scraped reference dumps and the LaTeX template
|
|
59
|
+
trees are not copied, and every drop is named in
|
|
60
|
+
skills/ai-research/SOURCE. `bin/tezgah-import-ai-research
|
|
61
|
+
--check` verifies the tree against its manifest.
|
|
62
|
+
|
|
63
|
+
skills/hallmark - MIT, Hallmark contributors, made by Together AI,
|
|
64
|
+
https://github.com/Nutlope/hallmark. VENDORED at revision
|
|
65
|
+
7ec868225c6caa040558c321ee0fefa0e73f83a8 (2026-10-22): the
|
|
66
|
+
whole `skills/hallmark/` tree - one `SKILL.md` (the workflow
|
|
67
|
+
itself, listed in its manifest because tezgah ships no entry
|
|
68
|
+
point of its own for it), 98 reference pages including 48
|
|
69
|
+
themes with their pictures and 20 macrostructures, and 10
|
|
70
|
+
zero-dependency Node scripts. Byte-for-byte upstream;
|
|
71
|
+
nothing is adapted, and upstream's own LICENSE sits at
|
|
72
|
+
`skills/hallmark/LICENSE`. `skills/hallmark/SOURCE` holds
|
|
73
|
+
the sha256 of every file and the checkout command; its own
|
|
74
|
+
test (`tests/test_skills.py::HallmarkVendor`) fails on a
|
|
75
|
+
body that drifted, a listed file that is missing, or a file
|
|
76
|
+
beside them the table does not list.
|
|
77
|
+
The tier map `scripts/sloplint.mjs` grades by is
|
|
78
|
+
`references/gates.md` itself, which wins on disagreement.
|
|
79
|
+
|
|
80
|
+
skills/design-library (inclusive-design-skills) - MIT, MC Dean,
|
|
81
|
+
https://github.com/Owl-Listener/inclusive-design-skills.
|
|
82
|
+
VENDORED at revision
|
|
83
|
+
6e0740f04b2130af60bc57abe3401b91e460e70d (2026-06-09): three
|
|
84
|
+
of its plugins - cognitive-accessibility,
|
|
85
|
+
adaptive-interfaces, inclusive-interaction - every skill in
|
|
86
|
+
them, 30 `SKILL.md`, at the upstream path, each renamed
|
|
87
|
+
`ENTRY.md` (bytes unchanged) so a host that lists every
|
|
88
|
+
`SKILL.md` below its skills directory does not list it.
|
|
89
|
+
ADAPTED, one body -
|
|
90
|
+
cognitive-accessibility/skills/adaptive-personalisation/ENTRY.md
|
|
91
|
+
carries `name: contextual-help-design` upstream, which is
|
|
92
|
+
the sibling `contextual-help-design` skill's name and not
|
|
93
|
+
its own directory; this copy's `name` matches its
|
|
94
|
+
directory, because a host that loads a skill from its
|
|
95
|
+
directory skips one whose name disagrees with it.
|
|
96
|
+
|
|
97
|
+
skills/design-library (designer-skills) - MIT, MC Dean,
|
|
98
|
+
https://github.com/Owl-Listener/designer-skills. VENDORED at
|
|
99
|
+
revision 9a6930cf84a822eb458624bd11c61aac5bbdf224
|
|
100
|
+
(2026-09-05): the whole visual-critique plugin (7
|
|
101
|
+
`SKILL.md`, renamed `ENTRY.md` the same way) and two
|
|
102
|
+
skills of design-research
|
|
103
|
+
(`behavioural-analytics`, `qual-quant-triangulation`),
|
|
104
|
+
byte-for-byte upstream. Nothing in this half is adapted;
|
|
105
|
+
the 12 other design-research skills and the plugin
|
|
106
|
+
packaging surface are not copied.
|
|
107
|
+
|
|
108
|
+
ADAPTED across the library, one statement: the
|
|
109
|
+
inclusive-design-skills copy of
|
|
110
|
+
inclusive-interaction/skills/touch-target-design/ENTRY.md
|
|
111
|
+
states the touch minimum as `44x44 CSS pixels (WCAG 2.2
|
|
112
|
+
Level AA)`. WCAG 2.2 SC 2.5.8 Target Size (Minimum) is
|
|
113
|
+
Level AA at 24x24 CSS px and SC 2.5.5 Target Size
|
|
114
|
+
(Enhanced) is Level AAA at 44x44, and tezgah's own checker
|
|
115
|
+
enforces 24 (bin/tezgah-design, `TAP_TARGET`); the adopted
|
|
116
|
+
copy states 24x24 as the AA floor with its success
|
|
117
|
+
criterion and keeps 44x44 as the AAA floor and the common
|
|
118
|
+
platform guideline. Every other adopted byte is upstream's,
|
|
119
|
+
and `skills/design-library/SOURCE` records the per-file
|
|
120
|
+
sha256 and names both departures.
|
|
121
|
+
`skills/design-library/SKILL.md`, `INDEX.md` and `EVALS.md`
|
|
122
|
+
are tezgah's own files and are not part of the manifest.
|
|
123
|
+
|
|
124
|
+
skills/rl-env - Apache-2.0, Adithya S Kolavi and the FineEnvs authors,
|
|
125
|
+
https://github.com/adithya-s-k/FineEnvs. VENDORED at
|
|
126
|
+
revision 26ab9c60915005b6f536968ac0928482bb18c0f8
|
|
127
|
+
(2026-10-01): the five environment-authoring skills and
|
|
128
|
+
every reference file they ship, 14 files under
|
|
129
|
+
`skills/rl-env/<skill>/`, byte-for-byte upstream including
|
|
130
|
+
frontmatter, at their upstream path under `.claude/skills/`,
|
|
131
|
+
each entry's `SKILL.md` renamed `ENTRY.md` so a host that
|
|
132
|
+
lists every `SKILL.md` below its skills directory does not
|
|
133
|
+
list it.
|
|
134
|
+
Nothing is adapted and nothing is dropped. The entry point
|
|
135
|
+
`skills/rl-env/SKILL.md` is tezgah's own and is not part of
|
|
136
|
+
the manifest, and upstream's own install route
|
|
137
|
+
(`npx skills add adithya-s-k/FineEnvs`) is deliberately not
|
|
138
|
+
used: it writes into a host's skill directory, which is
|
|
139
|
+
outside tezgah's router, its install report and this record.
|
|
140
|
+
`skills/rl-env/SOURCE` carries the upstream paths and the
|
|
141
|
+
per-file sha256.
|
|
142
|
+
|
|
143
|
+
third_party/open-code-review - Apache-2.0, the open-code-review authors,
|
|
144
|
+
https://github.com/alibaba/open-code-review. VENDORED at
|
|
145
|
+
commit 375f64cd6f67a48a135da83cfcf59c52669a09cb: two prompts,
|
|
146
|
+
`prompts/main_task_system.md` and
|
|
147
|
+
`prompts/review_filter_task_system.md`, from upstream's
|
|
148
|
+
`internal/config/template/prompts/`. MODIFIED: each prompt
|
|
149
|
+
carries one added first line, the modification notice; the
|
|
150
|
+
rest of each file is unchanged. Upstream's LICENSE sits at
|
|
151
|
+
`third_party/open-code-review/LICENSE`, and
|
|
152
|
+
`third_party/open-code-review/README.md` records the source.
|
package/README.de.md
CHANGED
|
@@ -25,8 +25,10 @@
|
|
|
25
25
|
---
|
|
26
26
|
|
|
27
27
|
Tezgah rüstet jeden KI-Programmierassistenten, den Sie ausführen — **omp, den
|
|
28
|
-
primären Host**, plus Claude Code, Codex, Cursor, opencode
|
|
29
|
-
|
|
28
|
+
primären Host**, plus Claude Code, Codex, Cursor, opencode, DeepSeeks dsh, Qwen
|
|
29
|
+
Code, Factory Droid, Command Code, GitHub Copilot CLI, Gemini CLI, Antigravity
|
|
30
|
+
CLI, Goose und Amp — innerhalb der von Ihnen konfigurierten
|
|
31
|
+
Repository-Stammverzeichnisse mit einem
|
|
30
32
|
funktionierenden Vertrag aus. Sich selbst überlassen, driften die Assistenten:
|
|
31
33
|
Einer antwortet auf Türkisch, ein anderer auf Englisch; einer nutzt Grep, ein
|
|
32
34
|
anderer fragt einen Code-Graphen ab; einer meldet „fertig“, ohne einen Test
|
|
@@ -37,11 +39,14 @@ auf dieselbe Weise.
|
|
|
37
39
|
|
|
38
40
|
## Was tezgah durchsetzt und aufzeichnet
|
|
39
41
|
|
|
40
|
-
- **Ein Vertrag,
|
|
41
|
-
dsh
|
|
42
|
-
|
|
42
|
+
- **Ein Vertrag, vierzehn Hosts.** omp, Claude Code, Codex, Cursor, opencode,
|
|
43
|
+
dsh, Qwen Code, Factory Droid, Command Code, GitHub Copilot CLI, Gemini CLI,
|
|
44
|
+
Antigravity CLI, Goose und Amp sehen dieselben Regeln, denn jeder Host ist
|
|
45
|
+
ein dünner Adapter über einem gemeinsamen Kern — eine Regel einmal ändern,
|
|
46
|
+
alle übernehmen sie.
|
|
43
47
|
- **Eine Fertig-Meldung braucht eine bestandene Prüfung im Protokoll.** Ein
|
|
44
|
-
Gate aus
|
|
48
|
+
Gate aus 17 Verweigerungsregeln und einem einmaligen Hinweis pro Sitzung
|
|
49
|
+
stoppt die entschärfte Prüfung — `--no-verify`,
|
|
45
50
|
`|| true`, ein in `tail` gepipeter Test, ein mitten im Flug ergänzter Skip —
|
|
46
51
|
und die Stop-Regel verweigert einen Fertigstellungsanspruch, den das
|
|
47
52
|
Evidenzprotokoll der Sitzung nicht stützt. Sie verweigert den Zug; sie kann
|
|
@@ -49,7 +54,7 @@ auf dieselbe Weise.
|
|
|
49
54
|
- **Forschung kommt mit eigener Bibliothek.** Forschungsaufgaben laufen über
|
|
50
55
|
OpenResearch, mit einer integrierten Bibliothek aus 98 Upstream-Skills, die
|
|
51
56
|
Eintrag für Eintrag geladen wird, damit der Kontext klein bleibt.
|
|
52
|
-
- **Jede Regel hat einen Ausschalter.** Siebzehn
|
|
57
|
+
- **Jede Regel hat einen Ausschalter.** Siebzehn Regel-Schalter — dazu
|
|
53
58
|
Repo-Markierungen — entfernen den Regeltext aus der Sitzung; die Regel
|
|
54
59
|
stoppt damit wirklich, statt nur als aus zu erscheinen.
|
|
55
60
|
- **Eine feste Antwortform.** Antworten sind türkisch, sofern Sie keine andere Sprache wählen, und
|
|
@@ -81,8 +86,14 @@ das Modell in beiden Armen kaum schummelte:
|
|
|
81
86
|
- **Eine Schwachstelle im Anspruchsleser der Stop-Regel:** Er markierte
|
|
82
87
|
115/132 Läufe mit tezgah und 90/131 ohne, aber nur 10% bzw. 20% seiner
|
|
83
88
|
Markierungen waren echte Fertig-Meldungen. Mit tezgah antwortete der Agent
|
|
84
|
-
auf Türkisch, und der Leser
|
|
85
|
-
Teilkorrektur als Anspruch
|
|
89
|
+
auf Türkisch, und der Leser zählte `doğrulandı` (verifiziert) bei einer
|
|
90
|
+
Teilkorrektur als Anspruch, selbst neben einem `doğrulanmadı` (nicht
|
|
91
|
+
verifiziert) für den fehlschlagenden Test. Diese Form ist behoben (Commit
|
|
92
|
+
9cb4d1d; Recall auf der zurückgehaltenen Menge 0.857, unter der Schwelle
|
|
93
|
+
von 0.9 des Plans). Ein engerer Rest bleibt: Eine türkische Teilkorrektur
|
|
94
|
+
ohne Eingeständniswort (`hâlâ başarısız`, schlägt noch fehl; `kısmen`,
|
|
95
|
+
teilweise) liest sich weiter als fertig und verifiziert, während ihr
|
|
96
|
+
englisches Gegenstück nur als Verifikation gilt.
|
|
86
97
|
|
|
87
98
|
Der Lauf hatte Messmängel, darunter Agenten in beiden Armen, die die
|
|
88
99
|
Grundwahrheit des Benchmarks lesen konnten. Die Aggregate, die Zählungen pro
|
|
@@ -94,8 +105,12 @@ Aufgabe und jede Einschränkung stehen im
|
|
|
94
105
|
## Installation
|
|
95
106
|
|
|
96
107
|
Eine Zeile, auf macOS, Linux oder WSL (Python 3.10+, `curl`, `tar`). Sie lädt
|
|
97
|
-
die neueste Version herunter, prüft ihre sha256
|
|
98
|
-
|
|
108
|
+
die neueste Version herunter, prüft ihre sha256, rüstet jeden gefundenen Host
|
|
109
|
+
(Claude Code über dessen eigene `claude plugin`-CLI) und installiert die
|
|
110
|
+
fehlenden optionalen Werkzeuge, die tezgah selbst nutzt (orx;
|
|
111
|
+
`TEZGAH_NO_DEPS=1` überspringt sie). Die eigene CLI eines Hosts
|
|
112
|
+
(cursor-agent, pnpm, dsh) installiert es nie: Es gibt stattdessen den Befehl
|
|
113
|
+
aus oder führt ihn unter `tezgah-setup --install --host-deps` aus.
|
|
99
114
|
|
|
100
115
|
```bash
|
|
101
116
|
curl -fsSL https://raw.githubusercontent.com/r1z4x/tezgah/main/packaging/install.sh | sh
|
|
@@ -116,18 +131,24 @@ gerüstet ist.
|
|
|
116
131
|
`~/.omp/agent/RULES.md`). Claude Codes Hooks kommen in einem Plugin, das über
|
|
117
132
|
`claude plugin` registriert wird; ohne `claude`-CLI bleibt Claude ungerüstet,
|
|
118
133
|
und der Bericht sagt es.
|
|
119
|
-
- **Agent-Dateien
|
|
120
|
-
|
|
121
|
-
|
|
122
|
-
|
|
134
|
+
- **Agent-Dateien.** Die Installation schreibt tezgahs Subagent-Rollen nach
|
|
135
|
+
`~/.claude/agents/`, `$CODEX_HOME/agents/` (Standard `~/.codex/agents/`) und
|
|
136
|
+
`~/.omp/agent/agents/`. In den konfigurierten Wurzeln schreibt eine Sitzung
|
|
137
|
+
sie in die `.cursor/agents/` und `.opencode/agents/` des Repositorys und
|
|
138
|
+
trägt diese Verzeichnisse in die eigene `.git/info/exclude` des Klons ein
|
|
139
|
+
(`TEZGAH_NO_EXCLUDE=1` stoppt das).
|
|
123
140
|
- **Eine tägliche Update-Prüfung.** Die Statuszeile fragt höchstens einmal am
|
|
124
141
|
Tag nach einer neueren Version; `~/.config/tezgah/update-check-off` oder
|
|
125
142
|
`TEZGAH_UPDATE_CHECK=0` schaltet sie ab.
|
|
126
143
|
- **orx.** Die Installation lädt die OpenResearch-CLI, über die
|
|
127
144
|
Forschungsaufgaben laufen; `TEZGAH_NO_DEPS=1` überspringt sie.
|
|
128
|
-
- **Antwortsprache.**
|
|
129
|
-
|
|
130
|
-
|
|
145
|
+
- **Antwortsprache.** Eine Erstinstallation wählt Antworten in der Sprache
|
|
146
|
+
Ihrer Locale (`LANG` und verwandte Variablen), wenn tezgah dafür ein Paket
|
|
147
|
+
mitbringt (`tr`, `en`), sonst Englisch; eine bestehende Installation behält
|
|
148
|
+
ihre Sprache. `--language en` verlangt Englisch und `any` Ihre eigene
|
|
149
|
+
Sprache: `curl -fsSL … | sh -s -- --reply-lang en` (die ältere Schreibweise
|
|
150
|
+
des Installers für `--language`). Nur türkische Antworten prüft die
|
|
151
|
+
Stop-Regel.
|
|
131
152
|
|
|
132
153
|
Lieber der Coding-Assistent erledigen? Fügen Sie dies in omp, Claude Code,
|
|
133
154
|
Codex, Cursor oder opencode ein (der Prompt bleibt Englisch; der Assistent
|
|
@@ -157,7 +178,7 @@ Upgrades und die optionalen Integrationen stehen in der [Dokumentation](docs/REA
|
|
|
157
178
|
|
|
158
179
|
## Unterstützte Hosts
|
|
159
180
|
|
|
160
|
-
**omp** (der primäre Host, gegen den entwickelt und verifiziert wird), **Claude Code**, **Codex**, **Cursor**, **opencode**, **dsh** — dünne Adapter über einem gemeinsamen Kern.
|
|
181
|
+
**omp** (der primäre Host, gegen den entwickelt und verifiziert wird), **Claude Code**, **Codex**, **Cursor**, **opencode**, **dsh**, **Qwen Code**, **Factory Droid**, **Command Code**, **GitHub Copilot CLI**, **Gemini CLI**, **Antigravity CLI**, **Goose**, **Amp** — dünne Adapter über einem gemeinsamen Kern. Bei den letzten acht hat noch kein Lauf ihrer Binaries gezeigt, dass sie ein Deny befolgen, und die letzten sechs sind allein nach der Dokumentation ihrer Hersteller angebunden ([Hosts](docs/hosts.md), Englisch).
|
|
161
182
|
|
|
162
183
|
## Wo die Tiefe liegt
|
|
163
184
|
|
|
@@ -166,4 +187,4 @@ Gate, Evidence-Ledger, Host-Adapter, Konfiguration und Entwicklung liegen in [do
|
|
|
166
187
|
## Lizenz
|
|
167
188
|
|
|
168
189
|
MIT — siehe [LICENSE](LICENSE).
|
|
169
|
-
Vendorte Skills
|
|
190
|
+
Vendorte und adaptierte Skills und Texte Dritter behalten ihre eigenen Bedingungen (MIT oder Apache-2.0), verzeichnet in [NOTICE](NOTICE).
|
package/README.es.md
CHANGED
|
@@ -25,8 +25,10 @@
|
|
|
25
25
|
---
|
|
26
26
|
|
|
27
27
|
Tezgah arma a cada asistente de programación de IA que ejecutes — **omp, el
|
|
28
|
-
host principal**, más Claude Code, Codex, Cursor, opencode
|
|
29
|
-
DeepSeek
|
|
28
|
+
host principal**, más Claude Code, Codex, Cursor, opencode, el dsh de
|
|
29
|
+
DeepSeek, Qwen Code, Factory Droid, Command Code, GitHub Copilot CLI, Gemini
|
|
30
|
+
CLI, Antigravity CLI, Goose y Amp — con un solo contrato de trabajo, dentro de
|
|
31
|
+
las raíces de
|
|
30
32
|
repositorios que configures. Por sí solos, los asistentes divergen: uno
|
|
31
33
|
responde en turco mientras otro responde en inglés; uno usa grep donde otro
|
|
32
34
|
consulta un grafo de código; uno informa "hecho" sin ejecutar una prueba. Las
|
|
@@ -36,11 +38,14 @@ en un lugar llega a todos los hosts de la misma manera.
|
|
|
36
38
|
|
|
37
39
|
## Qué impone y registra tezgah
|
|
38
40
|
|
|
39
|
-
- **Un contrato,
|
|
40
|
-
|
|
41
|
-
|
|
41
|
+
- **Un contrato, catorce hosts.** omp, Claude Code, Codex, Cursor, opencode,
|
|
42
|
+
dsh, Qwen Code, Factory Droid, Command Code, GitHub Copilot CLI, Gemini CLI,
|
|
43
|
+
Antigravity CLI, Goose y Amp ven las mismas reglas, porque cada host es un
|
|
44
|
+
adaptador ligero sobre un único núcleo compartido — cambia una regla una vez
|
|
45
|
+
y todos la adoptan.
|
|
42
46
|
- **Declarar algo terminado exige una comprobación superada en el registro.**
|
|
43
|
-
Una compuerta de
|
|
47
|
+
Una compuerta de 17 reglas de rechazo y un aviso único por sesión detiene la
|
|
48
|
+
comprobación neutralizada —
|
|
44
49
|
`--no-verify`, `|| true`, una prueba canalizada a `tail`, un skip añadido a
|
|
45
50
|
mitad de vuelo — y la regla Stop rechaza una afirmación de finalización que
|
|
46
51
|
el registro de evidencias de la sesión no respalda. Rechaza el turno; no
|
|
@@ -49,7 +54,7 @@ en un lugar llega a todos los hosts de la misma manera.
|
|
|
49
54
|
dirigen a través de OpenResearch con una biblioteca incorporada de 98
|
|
50
55
|
habilidades upstream, cargada de una en una para que el contexto siga
|
|
51
56
|
siendo pequeño.
|
|
52
|
-
- **Cada regla tiene un interruptor de apagado.** Diecisiete interruptores — más
|
|
57
|
+
- **Cada regla tiene un interruptor de apagado.** Diecisiete interruptores de regla — más
|
|
53
58
|
las marcas por repositorio — eliminan el texto de la regla de la sesión, de
|
|
54
59
|
modo que la regla realmente se detiene en lugar de limitarse a figurar como
|
|
55
60
|
desactivada.
|
|
@@ -81,8 +86,14 @@ ninguno de los dos brazos:
|
|
|
81
86
|
- **Un punto débil del lector de afirmaciones de la regla Stop:** marcó
|
|
82
87
|
115/132 ejecuciones con tezgah y 90/131 sin él, pero solo el 10% y el 20% de
|
|
83
88
|
sus marcas eran declaraciones reales de terminado. Con tezgah el agente
|
|
84
|
-
respondió en turco, y el lector
|
|
85
|
-
arreglo parcial como afirmación
|
|
89
|
+
respondió en turco, y el lector contaba `doğrulandı` (verificado) sobre un
|
|
90
|
+
arreglo parcial como afirmación, incluso junto a un `doğrulanmadı` (no
|
|
91
|
+
verificado) para la prueba que falla. Esa forma está corregida (commit
|
|
92
|
+
9cb4d1d; recall en el conjunto reservado 0.857, por debajo del umbral de 0.9
|
|
93
|
+
del plan). Queda un residuo más estrecho: un arreglo parcial en turco
|
|
94
|
+
expresado sin palabra de admisión (`hâlâ başarısız`, aún falla; `kısmen`,
|
|
95
|
+
parcialmente) sigue leyéndose como terminado y verificado, mientras que su
|
|
96
|
+
gemelo en inglés se lee solo como verificación.
|
|
86
97
|
|
|
87
98
|
La ejecución tuvo fallos de instrumento, entre ellos agentes de ambos brazos
|
|
88
99
|
que podían leer la verdad de referencia del benchmark. Los agregados, los
|
|
@@ -94,8 +105,12 @@ recuentos por tarea y cada salvedad están en el
|
|
|
94
105
|
## Instalación
|
|
95
106
|
|
|
96
107
|
Una línea, en macOS, Linux o WSL (Python 3.10+, `curl`, `tar`). Descarga la
|
|
97
|
-
última versión, comprueba su sha256
|
|
98
|
-
Code a través de su propia CLI `claude plugin`)
|
|
108
|
+
última versión, comprueba su sha256, arma cada host que encuentra (Claude
|
|
109
|
+
Code a través de su propia CLI `claude plugin`) e instala las herramientas
|
|
110
|
+
opcionales que faltan y que tezgah mismo usa (orx; `TEZGAH_NO_DEPS=1` las
|
|
111
|
+
omite). Nunca instala la CLI propia de un host (cursor-agent, pnpm, dsh):
|
|
112
|
+
imprime el comando en su lugar, o lo ejecuta con
|
|
113
|
+
`tezgah-setup --install --host-deps`.
|
|
99
114
|
|
|
100
115
|
```bash
|
|
101
116
|
curl -fsSL https://raw.githubusercontent.com/r1z4x/tezgah/main/packaging/install.sh | sh
|
|
@@ -115,18 +130,23 @@ con un código distinto de cero cuando un host previsto no queda armado.
|
|
|
115
130
|
`~/.omp/agent/RULES.md`). Los hooks de Claude Code van en un plugin
|
|
116
131
|
registrado con `claude plugin`; sin la CLI `claude`, Claude queda sin armar y
|
|
117
132
|
el informe lo dice.
|
|
118
|
-
- **Archivos de agentes
|
|
119
|
-
|
|
120
|
-
|
|
133
|
+
- **Archivos de agentes.** La instalación escribe los roles de subagente de
|
|
134
|
+
tezgah en `~/.claude/agents/`, `$CODEX_HOME/agents/` (por defecto
|
|
135
|
+
`~/.codex/agents/`) y `~/.omp/agent/agents/`. Dentro de las raíces
|
|
136
|
+
configuradas, una sesión los escribe en `.cursor/agents/` y
|
|
137
|
+
`.opencode/agents/` del repositorio, y añade esos directorios al
|
|
121
138
|
`.git/info/exclude` propio del clon (`TEZGAH_NO_EXCLUDE=1` lo impide).
|
|
122
139
|
- **Una comprobación diaria de actualizaciones.** La línea de estado pregunta
|
|
123
140
|
por una versión nueva como mucho una vez al día;
|
|
124
141
|
`~/.config/tezgah/update-check-off` o `TEZGAH_UPDATE_CHECK=0` la desactiva.
|
|
125
142
|
- **orx.** La instalación descarga la CLI de OpenResearch por la que pasan las
|
|
126
143
|
tareas de investigación; `TEZGAH_NO_DEPS=1` la omite.
|
|
127
|
-
- **Idioma de las respuestas.**
|
|
128
|
-
|
|
129
|
-
|
|
144
|
+
- **Idioma de las respuestas.** Una primera instalación pide respuestas en el
|
|
145
|
+
idioma de tu locale (`LANG` y similares) cuando tezgah trae un paquete para
|
|
146
|
+
él (`tr`, `en`), si no en inglés; una instalación existente conserva su
|
|
147
|
+
idioma. `--language en` pide inglés y `any` tu propio idioma:
|
|
148
|
+
`curl -fsSL … | sh -s -- --reply-lang en` (la grafía antigua del instalador
|
|
149
|
+
para `--language`). La regla Stop solo comprueba las respuestas en turco.
|
|
130
150
|
|
|
131
151
|
¿Prefieres que lo haga tu asistente de programación? Pega esto en omp,
|
|
132
152
|
Claude Code, Codex, Cursor u opencode (el prompt queda en inglés; el
|
|
@@ -156,7 +176,7 @@ actualizaciones y las integraciones opcionales están en la [documentación](doc
|
|
|
156
176
|
|
|
157
177
|
## Hosts compatibles
|
|
158
178
|
|
|
159
|
-
**omp** (el host principal, contra el que se desarrolla y verifica), **Claude Code**, **Codex**, **Cursor**, **opencode**, **dsh** — adaptadores ligeros sobre un único núcleo compartido.
|
|
179
|
+
**omp** (el host principal, contra el que se desarrolla y verifica), **Claude Code**, **Codex**, **Cursor**, **opencode**, **dsh**, **Qwen Code**, **Factory Droid**, **Command Code**, **GitHub Copilot CLI**, **Gemini CLI**, **Antigravity CLI**, **Goose**, **Amp** — adaptadores ligeros sobre un único núcleo compartido. Ninguna ejecución de los binarios de los ocho últimos ha mostrado todavía que obedezcan un deny, y los seis últimos están conectados solo a partir de la documentación de sus fabricantes ([hosts](docs/hosts.md), en inglés).
|
|
160
180
|
|
|
161
181
|
## Dónde está la profundidad
|
|
162
182
|
|
|
@@ -165,4 +185,4 @@ La compuerta, el registro de evidencia, los adaptadores de host, la configuraci
|
|
|
165
185
|
## Licencia
|
|
166
186
|
|
|
167
187
|
MIT — ver [LICENSE](LICENSE).
|
|
168
|
-
Las habilidades
|
|
188
|
+
Las habilidades y textos de terceros incorporados y adaptados conservan sus propios términos (MIT o Apache-2.0), registrados en [NOTICE](NOTICE).
|