v1.34.0

What's Changed

• docs(claude-md): compress global guidelines from 15.2 KB to 12.1 KB [C12, Fable 5.1, medium] by @richkuo in #259
• feat(effort): let Fable 5.1 run at xhigh when the user stamps or asks for it [C30, Fable 5.1, low] by @richkuo in #260
• docs(#269): compress the cli-dispatch skill from 11032 to 9145 bytes [C6, Fable 5.1, low] by @richkuo in #288
• docs(#275): shrink work-on-issue-loop and defer owned rules [C6, Fable 5.1, low] by @richkuo in #289
• docs(#279): shrink the fable-validate-loop skill from 9439 to 6567 bytes [C6, Fable 5.1, low] by @richkuo in #291
• docs(#284): shrink the fable-orchestrate skill from 8090 to 6623 bytes [C6, Fable 5.1, low] by @richkuo in #292
• docs(#265): trim fix-pr-review SKILL.md from 13343 to 12396 bytes [C10, Fable 5.1, low] by @richkuo in #293
• docs(#267): shrink the complexity scoring reference [C6, Fable 5.1, low] by @richkuo in #294
• docs(#281): compress the fableplan skill from 8129 to 6571 bytes [C6, Fable 5.1, low] by @richkuo in #295
• docs(#273): shrink the fix-pr-review-loop skill and keep every stop rule [C6, Fable 5.1, low] by @richkuo in #296
• docs(#283): shrink the milestoneplan skill from 8067 to 5930 bytes [C6, Fable 5.1, low] by @richkuo in #297
• docs(#276): bring prd-to-issues up to date and shrink it [C6, Fable 5.1, low] by @richkuo in #298
• docs(#280): shrink the sync-docs-release bot prompt in both templates [C10, Fable 5.1, low] by @richkuo in #299
• docs(#266): shrink the execution-plan-review skill from 11518 to 9040 bytes [C6, Fable 5.1, low] by @richkuo in #300
• docs(#268): shrink the codex-workflow template README [C4, Fable 5.1, low] by @richkuo in #301
• docs(#271): compress the pr-review skill from 10855 to 10433 bytes [C6, Fable 5.1, low] by @richkuo in #303
• docs(#278): shrink the pr-review worked example from 9440 to 4803 bytes [C4, Fable 5.1, low] by @richkuo in #305
• docs(#270): compress the pr-review-format bot prompt in both templates [C10, Fable 5.1, low] by @richkuo in #307
• docs(#286): compress the sync-docs skill from 7327 to 6197 bytes [C6, Fable 5.1, low] by @richkuo in #308
• docs(#285): shrink and refresh the claude-workflow template README [C4, Fable 5.1, low] by @richkuo in #309
• docs(#287): shrink the rereview-routing reference to 5957 bytes [C6, Fable 5.1, low] by @richkuo in #310
• docs(#274): shrink the work-on-issue skill from 10135 to 9545 bytes [C6, Fable 5.1, low] by @richkuo in #290
• docs(#282): bring fable-advisor up to date and shrink it to 7300 bytes [C6, Fable 5.1, low] by @richkuo in #302
• docs(#277): shrink the new-issue skill from 9495 to 5320 bytes [C6, Fable 5.1, low] by @richkuo in #304
• docs(#272): shrink the validate-issue skill and drop text its companions own [C6, Fable 5.1, low] by @richkuo in #306
• docs(#262): shrink the README from 26352 to 13284 bytes and add a size cap test [C29, Fable 5.1, medium] by @richkuo in #311
• docs(#261): shrink the contract inventory to 15765 bytes and cap it at 16000 [C35, Fable 5.1, medium] by @richkuo in #312
• docs(#264): shrink the milestone-workflow skill to 9920 bytes and add a size guard [C35, Fable 5.1, medium] by @richkuo in #313
• docs(#263): shrink both fix-pr bot prompts under a 10000-byte cap [C35, Fable 5.1, medium] by @richkuo in #314

Full Changelog: v1.33.0...v1.34.0

v1.33.0

What's Changed

• fix(installer): skip already-symlinked skills and workflows [C33, Sonnet 5, high] by @richkuo in #257
• feat(review-routing): post the opus review trigger with an explicit effort:high suffix [C18, Fable 5.1, high] by @richkuo in #258

Full Changelog: v1.32.0...v1.33.0

v1.32.0

What's Changed

• feat(milestone-workflow): add a targetBranch option across the issue pipeline [C45, Fable 5.1, high] by @richkuo in #253
• chore: remove remaining source comment [C5, Sonnet 5, high] by @richkuo in #254
• refactor(complexity): anchor every axis, report all five grades, and align bands to the score lattice [C66, Fable 5.1, high] by @richkuo in #256
• feat(fix-pr-review): merge re-review rule after a bare LGTM [C35, Fable 5.1, low] by @richkuo in #255

Full Changelog: v1.31.0...v1.32.0

v1.31.0

What's Changed

• docs(claude-md): ban emojis in every response and artifact [C10, Fable 5.1, low] by @richkuo in #250
• feat(milestone-pipeline): route stamped Codex CLI and Cursor CLI builds through a driver shim [C68, Fable 5.1, medium] by @richkuo in #251
• docs(milestone-workflow): merge serially on bare LGTM and resolve conflicts in-session [C20, Fable 5.1, high] by @richkuo in #252

Full Changelog: v1.30.0...v1.31.0

v1.30.0

What's Changed

• feat(execution-block): make validate and plan effort stampable per issue [C12, Fable 5.1, low] by @richkuo in #248
• docs(readme): sync milestone-workflow stamped effort row [C8, Composer, high] by @richkuo in #249

Full Changelog: v1.29.0...v1.30.0

v1.29.0

What's Changed

• chore(complexity): raise fableplan cutoff to score 71+ [C16, Composer 2.5, high] by @richkuo in #212
• docs(#195): apply the broken-test check on every code-changing path [C66, Fable 5, high] by @richkuo in #213
• chore(#216): pin every workflow action to a verified SHA at its current major [C68, Opus 5, high] by @richkuo in #217
• feat(#196): give a review finding a stated trigger before it blocks [C66, Opus 5, high] by @richkuo in #218
• refactor(#197): move the first-review boundaries and generalize the step-down ladder [C72, Opus 5, high] by @richkuo in #219
• docs(#198): state the measured discovery boundary and drop the dead MultiEdit denial [C93, Opus 5, xhigh, fableplan] by @richkuo in #220
• chore(#222): drop MultiEdit from the claude-run.yml allowlists [C6, Fable 5, low] by @richkuo in #223
• test(#221): guard behavior and shared rules, drop wording pins [C68, Fable 5, high] by @richkuo in #225
• refactor(#224): single Claude review path via shared engine [C68, Composer 2.5, xhigh] by @richkuo in #226
• feat(wans): add /wans next-steps skill [C12, Opus 5, high] by @richkuo in #227
• refactor(#201): shrink work-on-issue skill to unique gates [C66, Fable 5, low] by @richkuo in #228
• refactor(#209): shrink the validate-issue core under 8,000 bytes [C66, Fable 5, high] by @richkuo in #229
• refactor(#203): shrink the fix-pr-review skill below its size caps [C66, Fable 5, high] by @richkuo in #230
• chore: rename Fable 5 to Fable 5.1 throughout [C20, Opus 5, high] by @richkuo in #231
• refactor(#202): shrink the milestone-workflow skill to its unique procedures [C62, Fable 5.1, medium] by @richkuo in #232
• refactor(work-on-issue): tighten the skill and add three procedural gates [C8, Fable 5.1, high] by @richkuo in #233
• docs(claude-md): tighten the global guidelines and drop 660 bytes [C35, Fable 5.1, high] by @richkuo in #234
• refactor(fix-pr-review): tighten the skill and shrink its files [C64, Fable 5.1, high] by @richkuo in #235
• refactor(skills): shrink and tighten the fableplan and issueplan skills [C64, Fable 5.1, high] by @richkuo in #236
• refactor(validate-issue): tighten the baseline rule and shrink the skill [C64, Fable 5.1, high] by @richkuo in #237
• docs(claude-md): cut repo-specific and dead rules from the global guidelines [C35, Fable 5.1, high] by @richkuo in #238
• refactor(#208): trim fableplan under the 8,000-byte cap [C37, Fable 5.1, medium] by @richkuo in #239
• refactor(#206): trim sync-docs under the 8,000-byte cap [C37, Fable 5.1, high] by @richkuo in #240
• refactor(#207): trim the pr-review contract copies to short normative lines [C66, Fable 5.1, high] by @richkuo in #241
• refactor(prompts): restructure and trim the pr-review-format prompts and the github-issue-format skill [C12, Fable 5.1, high] by @richkuo in #242
• refactor(#210): trim fable-orchestrate under the 8,000-byte cap [C37, Fable 5.1, high] by @richkuo in #243
• refactor(fable-dispatch): trim the skill and reconcile the retry rule with the ladder [C33, Fable 5.1, high] by @richkuo in #244
• refactor(#204): trim fix-pr-review-loop to its unique stop rules [C66, Fable 5.1, medium] by @richkuo in #245
• refactor(#205): trim fable-advisor under the 8,000-byte cap [C37, Fable 5.1, high] by @richkuo in #246
• docs: sync README on work-on-issue gates and pr-review blocking test [C12, Composer 2.5, high] by @richkuo in #247

Full Changelog: v1.28.1...v1.29.0

v1.28.1

What's Changed

• chore: strip all source comments and ban new ones [C0, Composer 2.5, high] by @richkuo in #211

Full Changelog: v1.28.0...v1.28.1

v1.28.0

What's Changed

• feat(review-bands): add a sonnet first-review band and move the boundaries [C38, Opus 5, high] by @richkuo in #177
• feat(#179): anchor every review citation to the pull request head commit [C8, Opus 5, high] by @richkuo in #182
• docs(#180): add a worked example review to the pr-review skill [C35, Opus 5, high] by @richkuo in #183
• refactor(#181): split the source-verification rule into a decision table [C60, Opus 5, high] by @richkuo in #184
• feat(#178): make a review read the prior cycles before it drafts [C37, Opus 5, high] by @richkuo in #185
• refactor(pr-review): sign review comments with Reviewed [C12, Opus 5, high] by @richkuo in #187
• docs(claude-md): let the optimal solution set the tests [C22, Opus 5, high] by @richkuo in #194
• docs(work-on-issue): require stale existing tests to be updated [C6, Opus 5, high] by @richkuo in #188
• fix(#186): stage the pull request head and gate the standalone Claude review [C62, Opus 5, high, fableplan] by @richkuo in #189
• feat(pr-review): gate review remedies on scope and brake a diverging loop [C66, Opus 5, high] by @richkuo in #193
• fix(#191): classify pull request content as untrusted data in every review contract copy [C62, Opus 5, high] by @richkuo in #199
• docs: sync README on the fix-pr-review-loop diverging-loop brake by @richkuo in #200

Full Changelog: v1.27.0...v1.28.0

v1.27.0

What's Changed

• feat(skills): lead the validate verdict with the recommended next step [C12, Opus 5, high] by @richkuo in #174
• feat(pr-review): route review triggers by complexity band [C37, Opus 5, high] by @richkuo in #175

Full Changelog: v1.26.0...v1.27.0

v1.26.0

What's Changed

• feat(skills): require numbered plan steps with verify points [C8, Fable 5, medium] by @richkuo in #173

Full Changelog: v1.25.0...v1.26.0