From 352a9af285585f10554aa483f40af73b85e92d02 Mon Sep 17 00:00:00 2001
From: vitya <vitya.kuznetsov@gmail.com>
Date: Wed, 12 Aug 2026 17:10:31 +0300
Subject: [PATCH] =?UTF-8?q?feat:=20diagnosing-bugs=20v0.1.0=20(mattpocock?=
 =?UTF-8?q?=20core=20+=20superpowers=20Iron=20Law)=20+=20writing-skills=20?=
 =?UTF-8?q?v0.1.0=20(TDD-for-skills,=20anti-sproul)=20=E2=80=94=20=D1=81?=
 =?UTF-8?q?=D1=83=D0=BF=D0=B5=D1=80=D0=BF=D0=B0=D1=83=D1=8D=D1=80=D1=81-?=
 =?UTF-8?q?=D0=B4=D1=8B=D1=80=D1=8B=202/3=20=D0=B7=D0=B0=D0=BA=D1=80=D1=8B?=
 =?UTF-8?q?=D1=82=D1=8B?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 README.md                       |   2 +
 README.ru.md                    |   2 +
 dist/diagnosing-bugs.skill      | Bin 0 -> 6135 bytes
 dist/writing-skills.skill       | Bin 0 -> 4386 bytes
 skills/diagnosing-bugs/SKILL.md | 298 ++++++++++++++++++++++++++++++++
 skills/writing-skills/SKILL.md  | 208 ++++++++++++++++++++++
 6 files changed, 510 insertions(+)
 create mode 100644 dist/diagnosing-bugs.skill
 create mode 100644 dist/writing-skills.skill
 create mode 100644 skills/diagnosing-bugs/SKILL.md
 create mode 100644 skills/writing-skills/SKILL.md

diff --git a/README.md b/README.md
index 18686eb..7a68164 100644
--- a/README.md
+++ b/README.md
@@ -115,6 +115,8 @@ an explicit `adapted-from` marker in its frontmatter.
 | `find-skills` | `adapted-from: vercel-labs/skills` (MIT) — vendored copy, upstream pin TBD |
 | `grilling` | `adapted-from: mattpocock/skills @ 84fdeffd` (MIT) — family collapsed to one skill (pi hides `disable-model-invocation` wrappers) |
 | `brainstorming` | `adapted-from: obra/superpowers @ 6.2.0` (MIT) — divergent phase, visual-companion dropped |
+| `diagnosing-bugs` | `adapted-from: mattpocock/skills @ 84fdeffd` (MIT) + superpowers 6.2.0 concepts (Iron Law, red flags) |
+| `writing-skills` | `adapted-from: obra/superpowers @ 6.2.0` (MIT) — TDD-for-skills core + ideya 8 self-skill-authoring |
 | all other `skills/*` | `author: ours` |
 
 Adaptation policy: a clone is rewritten to our conventions (`.tasks/` boards,
diff --git a/README.ru.md b/README.ru.md
index b42558f..943bf6d 100644
--- a/README.ru.md
+++ b/README.ru.md
@@ -85,6 +85,8 @@ bash scripts/build.sh caveman          # один
 | `find-skills` | `adapted-from: vercel-labs/skills` (MIT) — вендорная копия, пин апстрима TBD |
 | `grilling` | `adapted-from: mattpocock/skills @ 84fdeffd` (MIT) — семейство схлопнуто в один скил (pi прячет `disable-model-invocation` обёртки) |
 | `brainstorming` | `adapted-from: obra/superpowers @ 6.2.0` (MIT) — расходящаяся фаза, visual-companion выброшен |
+| `diagnosing-bugs` | `adapted-from: mattpocock/skills @ 84fdeffd` (MIT) + superpowers 6.2.0 (Iron Law, red flags) |
+| `writing-skills` | `adapted-from: obra/superpowers @ 6.2.0` (MIT) — TDD-for-skills ядро + идея 8 self-skill-authoring |
 | остальные `skills/*` | `author: ours` |
 
 Политика адаптации: клон переписывается под наши конвенции (доски `.tasks/`,
diff --git a/dist/diagnosing-bugs.skill b/dist/diagnosing-bugs.skill
new file mode 100644
index 0000000000000000000000000000000000000000..48412d7a22cdaa93c9ddc580354b6a5d9b8f8695
GIT binary patch
literal 6135
zcmZ{oMNk|Jl!OO&cXxMp3+@&)xH}93gA-hW6MS%YLU1QI1PLD82MF#C`|oWp+kNP&
zu71~d__S2v;PC+f05V`F2Hmt(!a>-U2mrX}0RV{qYi%4Y?cLl#j&AlGR^Ij?E?re+
zH8oCG8*gJ%5Sb!4sK9&{A@esqdPthhQ~0%$?Dw?cq11MI{IwO&*o16EZm0!fem)I)
zCe`Uib!jWBrgHUlPNw)-obI^3VER<~#1q_t`6bu5$iqE4Vp=T`PXFZD+QZ&D$o;`X
zINr0%1Bp@jjGZemXC~^qKa>vg*Y+TFr<t7)$<DII$!JJEu;s8}j{OG8HTiO(h9{KA
z=yR@dA?oA)(Ss%F%ha*Qc%iYepZ9Rl?c;u3zt-3Hb@^0WgyKSy8<;ZcX5=TEQl6Ct
zo>^8k;EYRWW4YXz-I(o_q%hs9X&%Fxd_S6YkW$G_`t4;jZh%E+ofzJ^X(QjrKVQrb
z`ob91Nd{p;47)lQuu|9UG&&YPcuMg*&oi1bb*t{+Yw=+y_3OGXCsNz?59k?^!Om9-
za<`!F^7>sI@vqlVI1n;3S;}5-(rYmF+bVJ<Ko?^7S}^z1ioty)(3SO&ZIJ3@E?YwP
z@cg7K&eDhv2nxvCZIp1xvr|5-;%_MYc>t#2y6Xi)D=ywq8Z0!Y*?(KH)bq~YI!8++
z{Pq$IAI6+a>g#(-f=F$?dfZ$HeSmHbGq?I)r>3aPgCDNjgg)-}zG6L|eC)3Ebv+GE
z6_~v*Z4YAkzD7f&x?eGZM|7m#-w%8qUkijjqAq#|YyP_pxVa7ppTU1Rgh**mw)zQz
z2e%yWGO${AMi&Dy?$nKCz(2IyZ4xI2UW)QMx>H>bBCEd9hNoBfuGFDDVo7!KCLAx4
z_sZU~Z%+;-71LM=`j;dC40q@dUA$t5?}42|v~FV53ea6v)0~5K|3mc@d*=i|JFV(o
z^S~jOvw7M40z(P&^X1jf)79nP;MMrx(J|l6(QVUrw43AO$E&78YtZ}Y%3pE{$v$D8
z3eAt_eXGE4NSXx_!`6pHdnqD$Io{gVXU>^@dfI2U(Czc1_!@|q-DnPL-W1!ZAjt^W
zk%?r%n=||Gqe`)#4>he7<WQa<_bjZY1nWU`RlY!Qjah{>jD~{f5;KF;0tS-xyTo2|
zl&K&=F>ZhEv{d&ImaDY)x{F(M(uBl$cXEZ-xSvXp4-I)tjfcRo?Du+0WWl9to`_WP
z{<lNSPD;vx!Im5)!Z_`dHWea0c?v#7iwU7k?nBN-Ac4HKTpZ422(bI~&rF=&iJ^^}
zDaVP*JW*6tG8zi}AQv`(>Hb9~15UIy!3T(PR^fPD6KoFas8^oOmr0dhKu$%OpqWTz
zrqhSdo0*-kft)fRSCY~=Iaw}}I)AV{+rm*4!nF`)NK~$5iDr&NGC+A{VN578Ws{r>
zGT3#7+#D{h>z`#P%+9i{TFWwcwR#H(fga^k?`k?jI&qWIX8P~&jUp?qZ1?Al-1xGt
zn}NvcfLSLFJED@I19=I^e2Kc?#C*&<%xuD!5~I34+^Qi7Z%w=m{sQAsH$`jwyO-Ur
zg*YYP4l<%t5fMJRDI&>Z(2f-`eL|6k5DxXvt<1uUhmx{8vp;i3OZ0l-^~LfjVgduT
zAtDaLv_f3Sf?TWi%zndKIpu!1vGeTrxBY555ePt#1@y!=)Le@$3f}Ul>X&LoW(&L+
z6+fX>);qfv<(8u3qEi9Bn$-&20x5RYX>lbEU<?aB+J@pSGFLlNykNNI$NBJ_rASC>
zsmJ&*`)fi@ELKd|98bAan4g_-K|M(k?$<x@RMQ*H<xFu7Q8VRc1LwD^WKi7^hHmw|
zMo2qu(=2~F;&hm$5ic`+=XP$f+=Bcp1?RI;qhU3kNh}3-Shj%@TNkw#_2^rM`9JYG
z#Em(A?(=#wc^{U`N=`of`q90s>{%W4Mw-D|mJ{(q|Jbe(*aLx*i~R0~aTYl<qDG}5
zu^N05t61<P^=lM0Re#Jp%hrQlSo;NX??jxQWG;bx7^-!kTjp556Mm2EPEv{kNgNXD
z+yuxy!A{IR+Z2u?W-tI+YsMKmTfd&3z{n6^7w|QBqB^Ge*BXIpv`Fy1O_Bl^Z(C-U
zlU|xZ$x&|@p0u4hzw6IFRuoag^H1La`crDYC{(@G;~4jobM;NIT0QpLA-QF^u|o4E
z_Xb!>lFY>Y>a8@6lpd4~1I7`;M|6bb@)tUpJdUxZJ9S!8@^K9LwV7HrDbVAvOMT7m
z8q(SaXLA}iFtM_i4yoA68*DRHm-2e5KzR0XFW5HsStG=y83ohE^1qi4QdT5JsKn_}
zJeW?#<US-N%U{|Gq+^zSgB{Q&A827RVKcQzA<*(61Yp&|{<1kgy|~bA!zD3g=T~d?
zlwQ?~W5Fo1QYVwYQ0)~!o)u8HcZ@UqQFhtnbTEcz5E(^+M~O0M>tooGg&A8HQt(KJ
zt-s)A5!h%^{J@BiOae`Pn6!thgRoB;7XQOZ3&l^1Hg@N<zx=UF)D|-rDf}M7N=5e^
zN@$!)>YTT1oGyOUb>4@OM5`JlLj_8B-qxQ|oXu-BLh(|YCoyqCC@Y^Y2P74Xy`)N3
z>}jG`4vFt$7`zSn^Acgf<0UuSnUw@WA_2^Q@mramWQV|UL?%BlP+(Zc(P_@qJ6z`&
zKJ)8Sm9n2`>cKNFPSD?JdV*^{vDEr}It+gkisWRVL-@_Jg%WSBx|)T(7OW5~D^chL
z`=$UT35aS1_x#`*U@eLg8Tf2WAvgPn-14O2D6yV=9)!Brew}G|7;_X$UhUKk9hW#E
zSDX0hRsmZgf+{|SB|weX<Q$ViP_p4#v}?ifopPrW|2R5K0iN!O<P1K`IsVp^9mfxD
z2afsxMTKa8ghBjo7l$xIiL4cUe^07M2rP;0JxK><DyGxm5cKtUG#tQbQwY%gg*-6O
z>~&7I+)uI}9<sGYMH{HRRgv9eTm#e<#p^QXE2)m+FG{^$M&f?mfR=IC(hgvYL<{G(
zyCmFF@Ug(V9ufz>HjhPrxj?&NI`g8&{b5e6q5=j%EsVCaRmR52<JxulY_PdOTKgt_
zQoIEoc=YP&V56b@kC#K3KlSkC9qbItt;z$Bh`BC-5Y9@L&O!La==2-a^41yh5HB0d
zl9v~EOvegAH}3!!WXdctcR^CniGeQ?YFV%4(`QT>y-<b5gKvTjhyhCW{nQF{df|ox
zuSG>!@EH}^9@;$bmaqM|LJ~M&Oj)-|j@G1j-e?t?GH=Fd%EA(us3D$v%zu3}xL>?P
zh?g3P$g}ds>(d6MZOv(3?2yzohJ4ACSk(R@;Id?dq)WjlML=<Zx{OY?((#|;COEY!
zN<TJZeMFE9OXps6MLJQg&;;#dzKc%x%A+bm+ocWrO~$;%3HeAa-=lOpV^mQ|w3YOm
zs2uy42$H6h#{2kv1(Nv^$}xKqR+o}#`YYW>@zmd&PmRPEFMNzggFgz72{D*NEKcxo
zI;;N>t%Pjr6d^4{PFfFkJ?g0;GQs}@3h!X;x{N4za@Z2}Q3r_sVgm1L&kblVCU7&X
z*r<jbhA+)@g)4=w{&dB8eQkOAwAMKVmP6{-jgH(oBPHpc;%m;+HXAwRzi!yFH|agV
z8I3OqiGAzk_i<#f%OOxVpy@XJCdq_}tFvyHT*d%D1}i)a?Qs-8Pu{Y5?@=(Q3dUoe
zFr&1*hRujMq?H_V(k89xA#M?qI;YkD+lB8%ua#;AxB74HRt&XDt_4oNhJVI1j1wso
zXUP|#2QfaFS<j?vV=5=fWsQsN+@FI)tE67o%7IdRKg3Sy-{zGXp7UoEv4+{XLRv;0
zXW98u$f$Xl&&uZ!Y&C75sb2FVTm`{FXSQQ3+ouHu(L#&iy9lQG7lvyyF+e{xUhK;H
z_-MI=UY}66rg))jv@Of2@zLhImGd)hM*6)5=*q24q?ZjfR|wl~Q~zBJJ%*H!5j&(l
zUQDxwafm;!vl|zhkf_u>ArArUG0uO9?^ZWGK*Ml#Xc+vXoXjQ9AtybDJ)8)`{$fy@
zr`jiMIox^xfAn7U!>Rf{uKx$lWl5q&Li|EoQTm*TAoLYera+ISe_F{cL=V#tGmP=c
z`S;a!26Y${Tsw)mlAiU}pc5f|q*(*T_6(U-t7DUKTeO#uSZWdCV*eV+T;ZBDx{V}R
zgmc0sgK+Y0+EjcpPU56$g=luDB`k7Ld~wu|X;l`q%8$|nq@6@wc#iD6V9eZh)NnG~
z#A@W;TeMW9{-`7hWLe|N8VlATR#jPem{KZ5@@u64zGTJ;pibGk=^9p9U!kokA!=CP
z-}K?+c7smKquvsXtvt_$(<_)4iYOd~l8omj1`3VbL}9l(T{;cQl+MYX>%w`he~jMC
zq$7FgcdDv;`!sduFmvR{Ce6(C9f*(kh2kLHaaVIZUbZt0#{Me#vAC#S0lK&H41}tO
zqBY<bxtBLUp<5{PoOU{%LaN&ZcOhB6uB~uqMVb}qICo^}in|l<^_PiAcDsQreWftZ
zpREB7u~@%b3HK4jQ#nW0zaLQf9LeAEFm0>F1b=L5ZoTaV-FKc2lEsN)Sz`GA3sGIn
zFaz^^<19s>0*iC&rWlv*`Ds6w$0f6{R+P46E~w$z7j2=)meLinc)h(Z*ij<X6;ZG5
z>fm~0q^m0Fm4U+xqoFvO*41$HvgwbPP0<);zt`RRr9d|{715tAq`p?pH@BkDsxed<
zU`IChOb>pY4<dFhrn+)lI9Z_E>1ogfdWb3Lg2R>p26^yhl29603-z@I6HxC~pM$Ta
zhl)ZqdFq#}t5$16FF=(SCYd%$LaBwC0y&2bXEs8uU8gZ4mdp%P9P;<LE?0vYuWj<1
zK<v+y4K%}EU_*shW@@pzmsaN87jp3yc+PoFvO-=TaMfOnZ3+}HtSpK*mJDO3{QmnM
zm`dRAb1|J?3Gd)jo6(N#hu0_3jZ+E*@xWVyjmu!gO-WZ}a147)co{<_;U7g6|KDk%
zq4Z2qG}<;!l@P@+PfsqHb{L85ltjOLcoZE>YaeWA{^%T|NPb6NL2l3|IxcV0q5R<0
z_@@=Kf1&2AH-I53waW4^dsl|$#C$oQ-9cn8m)|6s{`+DS)hhdh-ucqwcFhH)xH#Vb
z4)Z=|0FQK!A^xYn>vrvw?Z-cvLWNp)6gg5&T=)d)Zf<LgcgYBxE!G_bc{BSYC7+I-
zMrB{7Z_kM`JtVHa>-MwGZMbc><gEfbQv4F_kVob)e$gKY5I@1qzws*PIe0@SSX|_^
zEGy}jzn%lYg<Pwtjl~Fh)i+ATa6o_T#%PF5bJD@%_;~0htSb_0lEs6DL){?2-GWmr
zE+H8~tSd7d`PVQeZZo`~Wm~3E;vu}F*Q6k?=I82xE$M&Vo>nW1c)YnZNtf7;eguf6
z?;(ip#?4Xm%9^x8V<<_B27%2WFQ?$=wXevvnO}kVbkR7`_RqRmEfZVpN9%m}L2s5^
z$BOvVzDhM_B2$&0Ox;^BVcIkBZCsQUlguNOFX5iR1c~>sC&&C30V;xWILVE5dc0kX
zhRFy`z|`-k5+tByW}4S;6tpcLw^!p&C_p)#aU{h;SNc!W_&KgEvt4LZ5AqL+m(y!)
zY*JY<<j*X_tBeS5!D}YS(P;AdvQ(C$IhPY29Iq5awk&j{_OraG$2jd${2S(vUev@g
z&Frlbc=kz@#!C!hrKtex1N!cmS*I67N~L3L+;lPlSqJSYGrD^C<s62Zupv}UxP2df
z&FBdv?>9l(4m(c1g^Z%M?i73BXJ27E`mvipngqyh%mD3S-hw8QSv^I)XXo3u^u-AA
z!fPC(P+c(VW+wkY@3qxhgnD>IQ;5p*kTkMyF%c_Nmof=AR9mzJYC`r0XU4Lj9koZ<
z;WTHQxQ*-1YWGo(5R>bho+NJ)Z_P4cu}5j#WOWo8wZxMS{mtp}QPrZgN#?VT>Nf0#
zytc(o4tRQDp-aC^DahfAtR$nu*LD*-*8mzO{WCssx^&$#5Y5H479ag8X;2%JC9bfC
z!G!jeMjqA|2A=~9c7BpwtfeMts!sXura*TAB%7T1HbPIy6~1Fwv2Mb@={zu#5T-&L
zWKzt3Ht4uHh1k7Af8gKHurx$SGI%=3vr~CF#<U2=e7^ur<wn2x#&rK#7Yos5YxJ6r
z+e)Dvv9dmK&(-f-ue0!$36!`cB}Hi}2!3ct^WX^vGg>e*xj4?6X#X8b$Ti37ujJSp
z{dPi~`TemY_k-DdT~>NQ^~qcF5%n3V1Lwza%YmDWSP1g52C19h=A6-zT|$IRmOLM^
zXqx05wb^!V%Y$WchdL8W{qCzyu_@x?6h@)Pr;po}M{lb^`eK@P9blU`$h14iAm2S}
zU~PjxwixuFm-i25<#Q7Du3DEICf8ckNE0KefU?bI!HlRZILoCXP@wk!Q7su5jOLU@
zq+y3KqS2j3ls||Rq4v!XmB!j^OA5YK@YSB5Gzj>l(vcR06|mAlf7s!&3TNoumKVXV
zSf$m)T75SM)ZxnXm%}5V#-@I7_<84HsVc#<5zq6Jn_}Y(-B9#hyIX#pP1apI#uY>`
zoiNbIg+=M-5JX8I%@U1y+YC3pv8*So9`dW|^Z6Yml2gQ;0n!AFzC4Bo?UWKgJJdff
z_T3Z2?gZ+it($3xi)Zemn1fLPZ(HUM)>Itf^>^83s|eoN=QCEKxvHEg+(3`&i{2{h
z<{jsif9-nO!Cn$h_aeOc;|z<Va99luh9HPTD(z?8Kog;+{FOsTGXgiCjs)S<c|uEd
zk@Sj9jKv|&j_GR>>GGKLXTyvv%sD0;&~8|BDIp?;uD(^b)fJQ%sI;B_7)>Sp=L-XG
z)UiX(F%?fU?o%Bm60^I<ZS$YmyA;W-67On<veP6zZ}M9A*Qb`LUczDRvnN-gL6tLO
zQyo$J=vpMDv(xDvY({B^8X|nOouSROeoG1PG7jG+(A|QVO)9gO0?E?hWwggb6g3(|
zERt6^_0ou@VY4O>re^keW~JRKCRk*`2eK=HOR2WzHUd3IYZ~3njUo8Nt=XyaGLov6
za3vVc_mhdj-Hwnl>Fh3dK9Kt&x)qDFwzN02vd5G-%V(%)Wbo9+vrfDAqAiTU<Z?%3
z4pt_RGbLvCHsLjt{(UJokm5$?Pn$1^)Mw`#*V{AhSLYAnUDdx|+!AqRbtqLa!k)%4
zThbqTkz|OjYc~1rEZv4{(0Cy@P?2KWC*jDetS<vvWl?MyKexvRicqHaT09W4nJ#~w
z8_&rawQ@q{O^}1Ue0d~AmIEO&7)jp>(C>;Z1rA9Bv!m+gM%p9PjyBxd@Jvoq)LNQr
zYFa0n$S-Mv2fSgg<z;9XeGTPd)!p$~%f1&m)(*b>Ns6=?&~eDP!al9uSHPIUJVPBe
zItQFYB)F)IqW&EF&iM8TRKeuT8*ybNH6cl>?w+&*=hZY&2r(eyK^AipPWQq>t`02u
zp6JMs=rNEqI7Uu=-}Uv_hW_j7kTmuS-%4>*#|wBK?Zq^vukFbp_{txbpSSN>BzQ@#
znqBUjDn>EXuq$@>COJZadl|&aw)E#$!Pd|UBXjfaiK?f=%07~?OzYz9stfjUnq`*o
z<{pSVl_pKw2uD(IK7Si@@Z*eQYXtgP)8PZSO%JiY)reu7L(082^&8<_vCd@anH*rQ
z#Un>4l1#Q>_6=<GZf#^s2nYCuZfZO}Qs%rG@wF5z?NFMIBF-TMel{St3H10-|D=m`
z+<E1L^}$nH+r4snPCzqMHLd4XIV0(rikn`1q^8keBEXdjOIio1*|c=xb3qE4+Hj{>
zq=}NU)dBu%S3|NWYSmb$MtwN8WjRLn3*e|#I~niK4q0C(&zzT-jbzx8a7c77nF_F+
ztk_}Ny1XaGNi+b<1bt6qXa90ySFq@R@m<K<Fg?jHOml>SXc-m*#U6W^B^6jPr;8t_
zwlAMj{l*{{oon9$tKX1qiJpE<_~Ur>l)H$l20poR8THf_#0*C&j6+N|Q5dgl#*>c;
zZ#XubWLtP<`1k>%r3wR!5BUG#{C|A@|LFfi^;)V3i2pl-`OgafCFDQ64*>iRk|m{4

literal 0
HcmV?d00001

diff --git a/dist/writing-skills.skill b/dist/writing-skills.skill
new file mode 100644
index 0000000000000000000000000000000000000000..10f88b7efc33dfdad2d38f5a8cb1ef9e683dae6e
GIT binary patch
literal 4386
zcmZ{o<y#XDpoK?`QW#1}$w*;z_vmo+=tjxWIvA2Nh6)IyOAsU`UD8OSG!i2vB&4Mq
zh5NhT?w5Pchx0t|bN+(Y5Jp5y3jhGf0A$hkYz^t53`rmWfS3yaVEB)sUbrLOy<G)g
zdboRfz7&STbaZuvyqw~!7ULKC{o6O;v_Zo7#hr~k><pQe)qvoL8w#Yu`J=9!QRQ_d
zPg&~_howYpwh^;xbwnh4`+!CQX@8p!y0dH0Y!wiM9*11&9Ot}@@zlpUJJq0F`b*)T
zM}MH&Ulwm8nds!C#r;j^=IK9_gnNA%){>iY_L|skd2yxNY+8g3foGK>&2h&&ygPkI
z=|`K0%RQv1arpJ^GJf?kuqD8Bw(5a-X7C`+<@1z1pHFT5*UTN<QVRud$x3RO-Ax6=
z)VW%LRU;kbWh)BL?X&OA)hapCaURNgtoSY^3l)6!owlCBb$q;i$@vYPEL8SmSa8^j
z<B*`RW2f6dzko`lrOR|-X~52El|imj+zIW3ak_+T@1;fCsiv)$-lwF&tt@K8sOBUh
zp20#hHu}M@H6OzRqB5VDt5DT7?B;l`-j5N+nE}s5_$8e;w*1TKQx`pTi{8XZPr2cg
z{+?flSKh5%hgU6!{@vU~?W{p<?pCfB&u*v>ctJNSjhFwhGyfqfSlDuS32<Ftl1;jk
zXgc_AolW9l!NG{_&`N^B-%*q1qsfILpcURa654<c_4LM%N{ZZvp1r#l6t!qh|02t3
zu6xEbIl-X&%pJT)T88>qc}irzjh(rR8)~~bywzq4{j246N4l`O*mnTEy}rIUUWIPk
z9aP=zUZWm||2wi*x&?apP91<(-clbtmlQ_GpE>`|^2D*_D9Fk~yQ0A+KG#p?veqRb
z$y-*g-USZ-nAu<~3Upd<Ovq_w5FN%w`>WQP-wl`g)OM5=qrCwZ5**zxBbCnclAA;q
z`Z3S?+1CTKPO|grkxWb&m3&Mt;$4K<QlO*&q}&@suOu2(S$^Va_4@-|U)^Ipd}?K~
zpe?t#hu0UkrtYO-dLX!dDxLlUvDjtD5$gVdWm?Iohgq)v1?R-~GGn#vPb}m^>&^{5
z6t;!h-|O`e*`m5)Z@>0uLmGv|!)TLJ+2FeA?}~fVGBe})c94Ri&nH*)SySFFa%*f8
zHrD5i#Xs?{Tk&71OtHP7t%`ditq4^a^Y(~VVF|x$y^^gBZkKrlQiLWunN{gCEa$Y@
z6gLNnK|_PC|613}G+Pw&e%@rDTwj>J{v~2S6&!Y4dUp8tO7ydoBDO^&F<q2YdpR4N
zGZ12N4kAc=Bv9>NE`+e1is^wV@)|`u;*alXN0rIalP+}+Yfq&7BKApaT5M%*H`dGK
z$oa<}zyJ8Gk|p%q63UIojrY%g!i0?pbM-xm*{HuBpF#Z`n$OQ=dHvpA;@u#C%kX*Z
zJs|Er?Qs2+sZfiTp4s=E@^{FZ8{$0O-U%>)>~_Na5q@#4Y`=Grmb9TXWzR{e>p&&q
z>}D;=dl;9Nw7*<c)_0(E3d<1lGZ;;+{i_+1{^ra<f7^5+zkyxO=Jj>6Pg#JR*p$u$
zhWGixD}uex(vD?Cm?W66@+^<AeB_6{UrY2<zG*j~f%i{1CY%&1nPg~jp?{IL$Ct$Z
zQA;Z_?~p8Y{aA2=?qX0JrSWtjPJ4LEkv4~JPXACX53>+yj7zdrRpsVHuJ}>EpqMBj
z06P50gQGz3b4Oc~Zh9QQ8(96en@bYi%N5|2O#Y0PVZ^3pyPLhdd2LFJ9z@2YV5Qm9
zo7$?=s@@Fgt*Iy)ghH)Pd!CqKA)Ei|Y=cX_08KA*Z&L#{<B$6^vQ^iupG#{rCP!Y;
zUF96Z9T+t;dyriZj|qJ-IU<kpSsT^>k-q5ld9I(`x3!|aFPl1CcV|Yyk(?0$@s9Q0
z2-QVEpa*3hXEmpTGN^K9j7`L#!#sZcZI_Vt7#)$VGgB4KOs-6ZK@JQ5J(W;dCzT|U
zy($UT=+BVu?}CO24*1r+2S<6^Niq<vPHxR|jjp46Ova2!kLD-N(v}r)QO*@hgD(TW
zlVyC3Tgu-!dMv<-W$=<kGx<!#_Mnkp*!bp7@?VsggsS`T%m2=xZ^Bm>nQ>0Zzdd{$
z1uXRJ{jecGD=EPF8lDaHdPU}3RuLe*ZrGN8JuqA;9OJ>@*jhZoQQW`p52YrFlU8iE
zob0?7km#=`Ih|K^BP7SKE6?2_?0aV#jwOd=eJG|<e~%(njpWcA^Td2M3>6j_4VK0<
zai6^0(phV{FC7!c^1ZkBZLQT6BGJvG=WL)6!aN+`tE7lmFqG#cl<|WGTd}I9vkLS%
z1+#u((1PWepcdOg-*!h;G?G#&cDNR0Nt@_c*P`*VSJA5C=$X8pjS-RKg%PkwmF|FI
zKafR>ln1d*y)$4^%{eC*Yo)v%8L<CrU}nSd<~&WO)w#Nqi7?45FcIq%8D(VE7HYki
z5EFbukei|Z>rO>$Le;60ji`Z!+fToKiW$M5rwPbWB|;;CUkTdM5Ully6N_1npN9s5
z%vI?I&+tJUHY{G*=Oz}vis%%tc7(<QISwjrA7(TJvT~2k6q_2??&dI2$u(5|^H_*e
z{}}|>ab5D1totsB6@z>32tGj<5rWH|*+ISAC=kUfGT1aEhU(SBnR8H%?#>e#f@81(
zxjFTMx=BO^SW1J1FDjw%PbJq!<DZzM*OwBIGfQO-cinzZH4#kZ{nx~_l=SZpZG_T|
zijfy3N1?uWf6OU=knu`qwy5p>aHa5bWN%kPVLhx}#=?GNIN$A<?hep4CLhG^-w?}}
zZUzTZSlE>ga+7S|d+%r|^{8mXWC2>F=_hM6gx}FyQRr}&`pCAp*O!iX=iq5rQ`l(p
za+fZP8n-c4Gb0>Zx+W@J|BK%D3KJ_$%4G~Cd8GE?K{14PGw%<2O@^(Q6Y)p2!^#`d
zAeHD2(yxYKRYmGZ=t9+LIw$!$P9z8E=x%G-BFJ+8?ya*pW^DG~*=LUN9sBQJR{G@I
z0i~^60j$Q&f@4AjymbXA$i9*abJqZSJeku6c}V*HEYK0desRImcW{9(9e@RDeTpgz
z(7h_5d6<TWmA~ZIw2v0#c6FghBwkY+`aKfeA3DYXSCnb7&zw*+gc@BKr9Cco1R5|b
zy-(}XP=)FR(Kc|hz`9M1u@$3w@hZyF+=R1k7Gw0P?WD0<eJLocuWWp`_*S_4$kh8j
z$c@l@i$#V>VLCd+XKYF?%DXnuND{H$iotiDuultJFrkZKG&=X6;v=-kz*-!a7E`;j
zFcxpXOqsQu$nuK{&S9ZF3Tirz7NXdn(^mf2b16FtlnL&a(Y-Y52jx5ji_>V=lC#IQ
z`RQw?Zr`VWLCsMCw0DsK<Ui1Q%UUBY{cDFI_6tLZ@bZ~6Q5vU{DL+*sCfd$YE^0V#
zjZ&+^AfjonU;K+)OXd3PBR2}FlS7V?;U}3*SWvOJk2bF2(W)O@%xQ;lqJxCNNuOYM
z-T3iwCED=FZ6!?eOJw7Y(U9OgDvxD>;v|g!pT32|(k3cmjer^!j!zU2l%U*FX<nG7
z&kytVFkgtxACdEY>P!By@jLI}=CtIbRr{c?;R_a!9hqe?i(3zI7eGzYUuW1X(xs*)
ztPG~aXxP_j(bdm=7Oi1R0C8QfnXe0wtiVjh5_!#?A%9uw|B?46@9-3GkHKY=6Cv;a
zR(2IXOwVms;C}F>fd}@Jhs@VBDN4PG*WYl+7@@kH^y8zvgTk`JfW(mq;cx^V*AXEr
zPAD=IUl=3O8HUjSN5*!0+IW3S(LN`uQAu;Pm5h3tL6Zl9v=AL)o~FJBR=-}#yR!LY
zW?#!&t*FsC%|qMfDpj(OGC&GSNfdbxn7(h?6VQ&ynkLrQ4a*I9fA6=t)dMOO*jw*@
zrgb^$(>QA@?Smc~8V0HD%^y$Xd?sW^h^X+SCd&e?s4qOlK62Fsc5+6J*1CFE3a$A^
zV1S^<#8^tJI|G6R953VO7Wk7l+ie|m$Phz?7z*Z-{ABN|QRwE@Pd+z&9*2KBpXzsN
zTHcNm5mxjFBVXG><K!%IGJW~58IixkX$y$d3_-C4*AJgk{O$ve^YSU~5mJV7tSDD1
zW_^1)`E3l_y$DFIa<o_C7oFJN0Lo+!=dp@$)_Z7LWHoU#{c3CRqI{q#YWeNy!;P!+
zC9fwIQWUe`JI(Y8a;5j`N-Mx_x5uhVx-~!lU0`@3i;cnEqM7=`US`7T>-5rbYLlSv
z2gvI#fyC0xQy&kla!t)<QqZa4$fU^|VRV^qklHjmR~X)cZ8gIHz>w-*^eSsB`Y3}x
z0ai(*Aeio^uHlL-$V+*~sS}m#Ngu;RI9u`CEc;=~+XYHEI7tq#cN0OUzy@QzP=lW}
z?i_gcy!t7{=~gVO$ui*J)<6m^f#Yj2z|4^HAPEF5#{QJKGyWNhPJm^*q<u<SyY7Ii
zE(0|0iTHk<?DXkhMf{)2H8$}Z&9h=2amQ7WU~gUpUxt8li*sF?`ZSmFqLB#00?2l9
zha(KGDKYgti$Oq)3+*~F&ZYaNP>wuGe&<ALMd@CrgXJH^j~w4x2-Mt|T@ZY{H=LIh
zR1yk8DA+`df?Xz^YPJ7hwY9#%<gAZ7_{g-EUA8&+@khs~Br2!59ltO!Xjs5RM<k|J
zH*!+HF(5Q|v_J1^@>F9si&~z%rG!vL(!G*=6@%Q+GyCa|i`63<XhKXUoXMVJe;QaS
znifCE1{K)dUNs;F4)uB7=nS#i7#AB@q<zXbVS<@60b88ml3d<RT%?%DXG<L{qEfjr
zUet}tQm+zDF0@2b&+*7f1A>io&m#jlC%2eZ)5s!deCFcnIm&eI(#uU!6&WCovOH$6
zreb;^=4A`t)OoDjtpG?-W9*AT3`vA+QL6ph?}-UXNeJ!Ck(qCe0Zqa$F^bE39oqZL
z<Jvlhg4JgyJ>xozGJ~gn;7uAMMVdjXAbyPEn9*(3O@!sPvM)vbXho+UGnOgKNz<I@
z<IoBznwly1@3!=vkk3{`NLE6-$VT$j2TGVi5dlef|JMB&(sEtqBL@zW3BqqL3Sgi~
z9B5#P%W<|dsXNX>z6twwH{s;CGf<#EpUhZVAiK!)x4yZ@^q;+HNWwUxhpEl&ap>fE
zli?oKj|=qY-`<Xx*&BqwwJhm2UuIkfodciFpOO*JI_uhR6>xUE3#a{4bGd~GbnPCl
zIp<iGrc`o8K9j+0Q6_zY9RISRDiBuzspfnh;tDjgW{_v(eD{GjPfm91vlzZIDFEj?
z;sU8mY{<!82)q&+eNYH~UVLz=5gg3nXIT0VQRvR^EN<&>AD(HqN8L<TC7ZO~V}l(Y
z+~ms|B}?9ISMWcRO`~!Mo3$S8{niNmQ(!(De!h0lkpj!zkuKOG36I-qFN(PFvfX)I
z+<GURGAZ$0Ju(0zM`o|-Eh@2^_HF-lNv!@B;Pq28y&)Yg8EvQ-<JT^KU9|D=oe3a(
zJB84b_z6l!evfkk>28(rJQEedSoP^eUTYca?9t(mUwfD?^TG=@;Jf$Up{Y!Ztw&z=
z0@Q>C#2G=1S(oEaHsTNN|0N#3R+K-gS!<B2GP+~*u?v;9ubX_z&>k)345Vva>p?a3
z{HR1-wSHK2=g)acf)mb(jDhar7d#fnI@EtMw{IQTB)(BGzte|!ehx2FfjfaqLz?3?
z1zvT7=@7^@2RQqK2Uls@qAL4)_O6zn`F6;Ro|PFP@2Cx71cbDJ|6gSM=Q93J{vWMj
W2qOXh?~CBSYW%l-|CtZ~;C}$^7eB`U

literal 0
HcmV?d00001

diff --git a/skills/diagnosing-bugs/SKILL.md b/skills/diagnosing-bugs/SKILL.md
new file mode 100644
index 0000000..9759352
--- /dev/null
+++ b/skills/diagnosing-bugs/SKILL.md
@@ -0,0 +1,298 @@
+---
+name: diagnosing-bugs
+adapted-from: mattpocock/skills @ 84fdeffd12f2ee307994d1eb6feb48173b6e0502 (MIT); concepts from obra/superpowers @ 6.2.0 (MIT)
+version: 0.1.0
+description: >
+  Diagnosis loop for hard bugs and performance regressions. Use when the user
+  says "diagnose"/"debug this", or reports something broken/throwing/failing/
+  slow, or any test failure / unexpected behavior / build failure / integration
+  issue — before proposing fixes. Triggers: «диагностируй», «почему падает»,
+  «разберись с багом», "debug this", "diagnose", "it's broken", "why is it
+  failing". Cross-agent — no tool refs beyond generic harness commands.
+---
+
+# Diagnosing Bugs
+
+A discipline for hard bugs. Skip phases only when explicitly justified.
+
+<HARD-GATE>
+NO FIXES WITHOUT ROOT CAUSE INVESTIGATION FIRST. If you haven't completed
+Phase 1 (a tight red-capable feedback loop), you cannot propose fixes.
+</HARD-GATE>
+
+When exploring the codebase, read `CONTEXT.md` (if it exists) to get a clear
+mental model of the relevant modules, and check ADRs in the area you're
+touching.
+
+## Redact
+
+This skill has you show commands, outputs and captured artifacts. **Redact
+every secret first** — write `<REDACTED>` in its place. Build loops against env
+vars, so the credential stays in the environment rather than in what you show.
+Captured artifacts carry auth headers: quote only the lines that carry the
+signal.
+
+If the redacted output is not enough to diagnose the bug, say so and ask the
+user.
+
+## Phase 1 — Build a feedback loop
+
+**This is the skill.** Everything else is mechanical. If you have a **tight**
+pass/fail signal for the bug — one that goes red on _this_ bug — you will find
+the cause; bisection, hypothesis-testing, and instrumentation all just consume
+it. If you don't have one, no amount of staring at code will save you.
+
+Spend disproportionate effort here. **Be aggressive. Be creative. Refuse to
+give up.**
+
+### Ways to construct one — try them in roughly this order
+
+1. **Failing test** at whatever seam reaches the bug — unit, integration, e2e.
+2. **Curl / HTTP script** against a running dev server.
+3. **CLI invocation** with a fixture input, diffing stdout against a known-good
+   snapshot.
+4. **Headless browser script** (Playwright / Puppeteer) — drives the UI,
+   asserts on DOM/console/network.
+5. **Replay a captured trace.** Save a real network request / payload / event
+   log to disk; replay it through the code path in isolation.
+6. **Throwaway harness.** Spin up a minimal subset of the system (one service,
+   mocked deps) that exercises the bug code path with a single function call.
+7. **Property / fuzz loop.** If the bug is "sometimes wrong output", run 1000
+   random inputs and look for the failure mode.
+8. **Bisection harness.** If the bug appeared between two known states
+   (commit, dataset, version), automate "boot at state X, check, repeat" so you
+   can `git bisect run` it.
+9. **Differential loop.** Run the same input through old-version vs new-version
+   (or two configs) and diff outputs.
+10. **HITL bash script.** Last resort. If a human must click, drive _them_
+    with a structured loop so the captured output feeds back to you.
+
+Build the right feedback loop, and the bug is 90% fixed.
+
+### Tighten the loop
+
+Treat the loop as a product. Once you have _a_ loop, **tighten** it:
+
+- Can I make it faster? (Cache setup, skip unrelated init, narrow the test
+  scope.)
+- Can I make the signal sharper? (Assert on the specific symptom, not "didn't
+  crash".)
+- Can I make it more deterministic? (Pin time, seed RNG, isolate filesystem,
+  freeze network.)
+
+A 30-second flaky loop is barely better than no loop; a 2-second deterministic
+one is tight — a debugging superpower.
+
+### Non-deterministic bugs
+
+The goal is not a clean repro but a **higher reproduction rate**. Loop the
+trigger 100×, parallelise, add stress, narrow timing windows, inject sleeps. A
+50%-flake bug is debuggable; 1% is not — keep raising the rate until it's
+debuggable.
+
+### When you genuinely cannot build a loop
+
+Stop and say so explicitly. List what you tried. Ask the user for: (a) access
+to whatever environment reproduces it, (b) a redacted captured artifact (HAR
+file, log dump, core dump, screen recording with timestamps), or (c) permission
+to add temporary production instrumentation. Do **not** proceed to hypothesise
+without a loop.
+
+### Completion criterion — a tight loop that goes red
+
+Phase 1 is done when the loop is **tight** and **red-capable**: you can name
+**one command** — a script path, a test invocation, a curl — that you have
+**already run at least once** (show the invocation and its output, redacted),
+and that is:
+
+- [ ] **Red-capable** — it drives the actual bug code path and asserts the
+      **user's exact symptom**, so it can go red on this bug and green once
+      fixed. Not "runs without erroring" — it must be able to _catch this
+      specific bug_.
+- [ ] **Deterministic** — same verdict every run (flaky bugs: a pinned, high
+      reproduction rate, per above).
+- [ ] **Fast** — seconds, not minutes.
+- [ ] **Agent-runnable** — you can run it unattended.
+
+If you catch yourself reading code to build a theory before this command
+exists, **stop — jumping straight to a hypothesis is the exact failure this
+skill prevents.** No red-capable command, no Phase 2.
+
+## Phase 2 — Reproduce + minimise
+
+Run the loop. Watch it go red — the bug appears.
+
+Confirm:
+
+- [ ] The loop produces the failure mode the **user** described — not a
+      different failure that happens to be nearby. Wrong bug = wrong fix.
+- [ ] The failure is reproducible across multiple runs (or, for
+      non-deterministic bugs, reproducible at a high enough rate to debug
+      against).
+- [ ] You have captured the exact symptom (error message, wrong output, slow
+      timing) so later phases can verify the fix actually addresses it.
+
+### Minimise
+
+Once it's red, shrink the repro to the **smallest scenario that still goes
+red**. Cut inputs, callers, config, data, and steps **one at a time**, re-running
+the loop after each cut — keep only what's load-bearing for the failure.
+
+Why bother: a minimal repro shrinks the hypothesis space in Phase 3 (fewer
+moving parts left to suspect) and becomes the clean regression test in Phase 5.
+
+Done when **every remaining element is load-bearing** — removing any one of
+them makes the loop go green.
+
+Do not proceed until you have reproduced **and** minimised.
+
+## Phase 3 — Hypothesise
+
+Generate **3–5 ranked hypotheses** before testing any of them.
+Single-hypothesis generation anchors on the first plausible idea.
+
+Each hypothesis must be **falsifiable**: state the prediction it makes.
+
+> Format: "If <X> is the cause, then <changing Y> will make the bug disappear
+> / <changing Z> will make it worse."
+
+If you cannot state the prediction, the hypothesis is a vibe — discard or
+sharpen it.
+
+**Show the ranked list to the user before testing.** They often have domain
+knowledge that re-ranks instantly ("we just deployed a change to #3"), or know
+hypotheses they've already ruled out. Cheap checkpoint, big time saver. Don't
+block on it — proceed with your ranking if the user is AFK.
+
+## Phase 4 — Instrument
+
+Each probe must map to a specific prediction from Phase 3. **Change one
+variable at a time.**
+
+Tool preference:
+
+1. **Debugger / REPL inspection** if the env supports it. One breakpoint beats
+   ten logs.
+2. **Targeted logs** at the boundaries that distinguish hypotheses.
+3. Never "log everything and grep".
+
+**Tag every debug log** with a unique prefix, e.g. `[DEBUG-a4f2]`. Cleanup at
+the end becomes a single grep. Untagged logs survive; tagged logs die.
+
+**Perf branch.** For performance regressions, logs are usually wrong. Instead:
+establish a baseline measurement (timing harness, `performance.now()`,
+profiler, query plan), then bisect. Measure first, fix second.
+
+**Multi-component systems:** when the failure path crosses components
+(CI → build → signing, API → service → database), before proposing fixes add
+diagnostic instrumentation at each component boundary — log what enters, what
+exits, and verify environment/config propagation at each layer. Run once to
+gather evidence showing WHERE it breaks, then investigate that component.
+
+## Phase 5 — Fix + regression test
+
+Write the regression test **before the fix** — but only if there is a
+**correct seam** for it.
+
+A correct seam is one where the test exercises the **real bug pattern** as it
+occurs at the call site. If the only available seam is too shallow
+(single-caller test when the bug needs multiple callers, unit test that can't
+replicate the chain that triggered the bug), a regression test there gives
+false confidence.
+
+**If no correct seam exists, that itself is the finding.** Note it. The
+codebase architecture is preventing the bug from being locked down. Flag this
+in the post-mortem.
+
+If a correct seam exists:
+
+1. Turn the minimised repro into a failing test at that seam.
+2. Watch it fail.
+3. Apply the fix.
+4. Watch it pass.
+5. Re-run the Phase 1 feedback loop against the original (un-minimised)
+   scenario.
+
+**One change at a time.** No "while I'm here" improvements, no bundled
+refactoring.
+
+### If the fix doesn't work
+
+- Count how many fixes you've tried.
+- If < 3: return to Phase 1, re-analyze with new information.
+- **If ≥ 3: STOP and question the architecture.** Each fix revealing new
+  shared state / coupling / problems in different places is the pattern of an
+  architectural problem, not a failed hypothesis. Discuss with the user before
+  attempting more fixes. This is NOT a failed hypothesis — this is a wrong
+  architecture.
+
+## Phase 6 — Cleanup + post-mortem
+
+Required before declaring done:
+
+- [ ] Original repro no longer reproduces (re-run the Phase 1 loop)
+- [ ] Regression test passes (or absence of seam is documented)
+- [ ] All `[DEBUG-...]` instrumentation removed (`grep` the prefix)
+- [ ] Throwaway prototypes deleted (or moved to a clearly-marked debug
+      location)
+- [ ] The hypothesis that turned out correct is stated in the commit / PR
+      message — so the next debugger learns
+
+**Then ask: what would have prevented this bug?** If the answer involves
+architectural change (no good test seam, tangled callers, hidden coupling)
+hand the specifics off to the project owner / architecture skill. Make the
+recommendation **after** the fix is in, not before — you have more information
+now than when you started.
+
+## Red Flags — STOP and return to Phase 1
+
+If you catch yourself thinking any of these, stop and go back:
+
+- "Quick fix for now, investigate later"
+- "Just try changing X and see if it works"
+- "Add multiple changes, run tests"
+- "Skip the test, I'll manually verify"
+- "It's probably X, let me fix that"
+- "I don't fully understand but this might work"
+- "Pattern says X but I'll adapt it differently"
+- Proposing solutions before tracing data flow
+- "One more fix attempt" (when already tried 2+)
+- Each fix reveals a new problem in a different place
+
+**All of these mean: STOP. Return to Phase 1.**
+
+## Common Rationalizations
+
+| Excuse | Reality |
+|--------|---------|
+| "Issue is simple, don't need process" | Simple issues have root causes too. Process is fast for simple bugs. |
+| "Emergency, no time for process" | Systematic debugging is FASTER than guess-and-check thrashing. |
+| "Just try this first, then investigate" | First fix sets the pattern. Do it right from the start. |
+| "I'll write test after confirming fix works" | Untested fixes don't stick. Test first proves it. |
+| "Multiple fixes at once saves time" | Can't isolate what worked. Causes new bugs. |
+| "I see the problem, let me fix it" | Seeing symptoms ≠ understanding root cause. |
+| "One more fix attempt" (after 2+ failures) | 3+ failures = architectural problem. Question the architecture, don't fix again. |
+
+## When Process Reveals "No Root Cause"
+
+If systematic investigation reveals the issue is truly environmental,
+timing-dependent, or external:
+
+1. You've completed the process.
+2. Document what you investigated.
+3. Implement appropriate handling (retry, timeout, error message).
+4. Add monitoring/logging for future investigation.
+
+**But:** 95% of "no root cause" cases are incomplete investigation.
+
+## Cross-agent applicability
+
+Pure methodology — no harness-specific tool references. Works on pi, Claude,
+or any agent. The sub-agent mention is a generic capability note; without
+sub-agent support the agent looks facts up directly.
+
+## Out of scope
+
+- Does NOT cover code review (that's a separate review process).
+- Does NOT write the regression-test policy (see `tdd-criteria` for the
+  bright-line rules on when tests are required).
diff --git a/skills/writing-skills/SKILL.md b/skills/writing-skills/SKILL.md
new file mode 100644
index 0000000..463ad19
--- /dev/null
+++ b/skills/writing-skills/SKILL.md
@@ -0,0 +1,208 @@
+---
+name: writing-skills
+adapted-from: obra/superpowers @ 6.2.0 (MIT) — TDD-for-skills core; ideya 8 self-skill-authoring (workshop record)
+version: 0.1.0
+description: >
+  Authoring agent skills TDD-style — RED-GREEN-REFACTOR applied to SKILL.md
+  documents. Use when creating a new skill, editing an existing one, or
+  verifying a skill works before deploying. Trigger (user): «напиши скил»,
+  «создай скил», "write a skill", "create a skill"; or self-authoring trigger
+  (Hermes-mode): the third time you do the same thing without instruction, or
+  the user says «запомни»/«зафиксируй»/«в следующий раз». Anti-sproul: BEFORE
+  writing a new skill, check the catalog for an existing cover.
+---
+
+# Writing Skills
+
+**Writing skills IS Test-Driven Development applied to process documentation.**
+
+You write test cases (pressure scenarios with subagents), watch them fail
+(baseline behavior without the skill), write the skill (documentation), watch
+tests pass (agents comply), and refactor (close loopholes).
+
+**Core principle:** If you didn't watch an agent fail without the skill, you
+don't know if the skill teaches the right thing.
+
+<HARD-GATE>
+NO SKILL WITHOUT A FAILING TEST FIRST. This applies to NEW skills AND edits to
+existing skills. Write the skill before testing? Delete it. Start over. No
+exceptions — not for "simple additions", not for "documentation updates".
+</HARD-GATE>
+
+## Anti-sproul guard (check before writing)
+
+At 45+ skills in the catalog, "just write another one" is harm, not help. On
+any self-authoring trigger, FIRST check for a double/coverage:
+
+1. Does a skill already exist that covers this? (catalog + adapted-from
+   sources: mattpocock/skills, obra/superpowers, vendor skills)
+2. Should this be a **new skill**, an **extension of an existing one**, a
+   **rule in CLAUDE.md**, or **nothing** (one-off coincidence)?
+
+Only proceed to the TDD cycle if the answer is genuinely "new skill". If the
+pattern is project-specific, it belongs in the project's `.agents/skills/`
+(copy, project owns it) — NOT the catalog. Cross-project value → promote to
+the sovereign catalog.
+
+## What is a Skill?
+
+A **skill** is a reference guide for proven techniques, patterns, or tools.
+Skills help future agents find and apply effective approaches.
+
+**Skills are:** reusable techniques, patterns, tools, reference guides.
+**Skills are NOT:** narratives about how you solved a problem once.
+
+## TDD Mapping for Skills
+
+| TDD Concept | Skill Creation |
+|---|---|
+| **Test case** | Pressure scenario with subagent |
+| **Production code** | Skill document (SKILL.md) |
+| **Test fails (RED)** | Agent violates rule without skill (baseline) |
+| **Test passes (GREEN)** | Agent complies with skill present |
+| **Refactor** | Close loopholes while maintaining compliance |
+| **Write test first** | Run baseline scenario BEFORE writing skill |
+| **Watch it fail** | Document exact rationalizations agent uses |
+| **Minimal code** | Write skill addressing those specific violations |
+| **Watch it pass** | Verify agent now complies |
+
+## RED — Write the failing test (baseline)
+
+Run a pressure scenario with a fresh-context subagent **WITHOUT the skill**.
+Document exact behavior:
+
+- What choices did they make?
+- What rationalizations did they use (verbatim)?
+- Which pressures triggered violations?
+
+This is "watch the test fail" — you must see what agents naturally do before
+writing the skill.
+
+**Pressure types for discipline skills** (combine 3+): time pressure, sunk
+cost, authority ("the user asked for it"), exhaustion/length, "it's simple".
+
+## GREEN — Write the minimal skill
+
+Write the skill that addresses those **specific** rationalizations. Don't add
+content for hypothetical cases.
+
+### Skill structure (our catalog conventions)
+
+```
+skills/<name>/SKILL.md
+```
+
+Frontmatter (YAML):
+
+- `name` — letters, numbers, hyphens only. Verb-first, active voice:
+  `pulling-before-work`, `diagnosing-bugs`, `writing-skills` (gerund works for
+  processes).
+- `description` — **when to use, NOT what it does.** Start with "Use when..." /
+  trigger phrases. Third person (injected into system prompt). NEVER summarize
+  the skill's process or workflow — agents follow the description instead of
+  reading the body.
+- `version` — semver; bump on every edit (project-discipline Rule 3).
+- `author: ours` or `adapted-from: <owner>/<repo> @ <pin> (license)` for
+  vendored/adapted copies — with the real upstream pin, not "TBD".
+
+Body:
+
+- Overview: core principle in 1-2 sentences.
+- When to use: bullet list with symptoms and triggers; when NOT to use.
+- Core pattern / Quick reference: table or bullets for scanning.
+- Common mistakes / rationalizations: table (excuse → reality).
+- Red flags: self-check list ("all of these mean STOP and start over").
+- Cross-agent applicability note (no harness-specific tool refs).
+- Out of scope: what this skill explicitly does NOT do.
+
+**Guidance form must match the failure type:**
+
+| Baseline failure | Right form | Wrong form |
+|---|---|---|
+| Skips/violates a rule under pressure | Prohibition + rationalization table + red flags | Soft guidance ("prefer...") |
+| Complies, but output has wrong shape | Positive recipe/contract: state what the output IS | Prohibition list |
+| Omits a required element | Structural: REQUIRED field in the template | Prose reminders |
+
+**No nuance clauses.** "Don't X unless it matters" reopens the negotiation —
+express a real exception as its own conditional on an observable predicate.
+
+### Micro-test wording before full scenarios
+
+Full pressure-scenario runs are the final gate but slow. Verify the wording
+first:
+
+1. One fresh-context sample per call; system prompt = the realistic context.
+2. Always include a **no-guidance control**. If the control doesn't exhibit
+   the failure, there is nothing to fix — stop.
+3. 5+ reps per variant. Single samples lie.
+4. Manually read every flagged match (template echoes masquerade as hits).
+5. Variance is a metric: five different interpretations across five reps means
+   the wording isn't binding — tighten the form.
+
+Micro-tests verify wording; they do not replace pressure scenarios for
+discipline skills.
+
+## REFACTOR — close loopholes
+
+- Agent found a new rationalization? Add an explicit counter.
+- Build the rationalization table from all test iterations.
+- Create the red-flags list.
+- Re-test until bulletproof.
+
+## Verification checklist (before declaring done)
+
+- [ ] Baseline failure documented (RED evidence: what the agent did without
+      the skill)
+- [ ] Skill addresses those specific failures (not hypotheticals)
+- [ ] Scenario passes WITH the skill (GREEN evidence)
+- [ ] Description = when to use only, no workflow summary
+- [ ] Frontmatter: name (verb-first, hyphens), description (triggers), version
+      (bumped), provenance (author/adapted-from with real pin)
+- [ ] Lint passes (catalog: `scripts/lint-skills.py`), dist rebuilt
+      (`scripts/build.sh`), installed (`scripts/install.sh <name>`)
+- [ ] README provenance table updated (catalog)
+- [ ] Cross-agent — no harness-specific tool references
+- [ ] One excellent example, not multi-language dilution
+
+## Deploying (catalog)
+
+1. Edit `skills/<name>/SKILL.md`.
+2. Lint: `python scripts/lint-skills.py`.
+3. Build: `bash scripts/build.sh` (→ `dist/<name>.skill`).
+4. Install: `bash scripts/install.sh <name>` (→ `~/.claude/skills/`).
+5. Update README provenance table (en+ru).
+6. Commit + push; bump version in the commit (project-discipline Rule 3).
+
+**STOP after each skill.** Don't batch-create skills without testing each one.
+Deploying untested skills = deploying untested code.
+
+## Common Rationalizations for Skipping Testing
+
+| Excuse | Reality |
+|--------|---------|
+| "Skill is obviously clear" | Clear to you ≠ clear to other agents. Test it. |
+| "It's just a reference" | References can have gaps, unclear sections. |
+| "Testing is overkill" | Untested skills have issues. Always. |
+| "I'll test if problems emerge" | Problems = agents can't use skill. Test BEFORE deploying. |
+| "Too tedious to test" | Testing is less tedious than debugging a bad skill in production. |
+| "I'm confident it's good" | Overconfidence guarantees issues. Test anyway. |
+| "Academic review is enough" | Reading ≠ using. Test application scenarios. |
+| "There's already a similar skill, close enough" | Similar ≠ cover. Verify the double actually covers the failure. |
+
+**All of these mean: test before deploying. No exceptions.**
+
+## Cross-agent applicability
+
+Pure authoring methodology — no harness-specific tool references. The
+"subagent" for pressure scenarios is any fresh-context agent run (pi -p,
+claude -p, codex exec, hermes headless). Local skills live in the agent's
+skills directory (`~/.claude/skills/`, `~/.hermes/skills/`, project
+`.agents/skills/`); the catalog deploy steps are our repo's convention.
+
+## Out of scope
+
+- Does NOT tell you WHAT skill to write — that's the anti-sproul guard + the
+  user's call.
+- Does NOT cover how to run the catalog pipeline beyond the deploy steps above
+  (see the repo's `scripts/` and README).
+- Does NOT enforce TDD for code — that's `tdd-criteria`.