Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
79 commits
Select commit Hold shift + click to select a range
eb8172f
feat(write-skill): preserve context-dependent outcomes
adamwhp Aug 14, 2026
a485d74
docs: add lifecycle skills implementation plan
adamwhp Aug 14, 2026
bdcc9f2
feat(elicit): expand domain-neutral elicitation
adamwhp Aug 14, 2026
dd74920
docs: add skills implementation checklist
adamwhp Aug 14, 2026
76036e6
feat(research): strengthen evidence reporting
adamwhp Aug 14, 2026
557920f
feat(prototype): add disposable visual experiments
adamwhp Aug 14, 2026
f96225a
feat(maintain-context): maintain ubiquitous language inline
adamwhp Aug 14, 2026
8df74dc
docs(context): remove duplicated skill definitions
adamwhp Aug 14, 2026
bc37d33
feat(modular-design): teach structural maintainability
adamwhp Aug 15, 2026
5aad60e
feat(tdd): enforce durable behavioural tests
adamwhp Aug 15, 2026
49dee50
feat(verify-change): add quality harness verification
adamwhp Aug 15, 2026
74db725
feat(code-review): add independent diagnostic review
adamwhp Aug 15, 2026
d7976a2
feat(maintain-agents): enforce lean repository guidance
adamwhp Aug 15, 2026
fe9f912
Reset implementation checklist
adamwhp Aug 15, 2026
1d8610f
feat(elicit): tighten shared understanding
adamwhp Aug 15, 2026
290823b
feat(research): tighten evidence handoff
adamwhp Aug 15, 2026
a1959bd
feat(prototype): enforce disposable visual experiments
adamwhp Aug 15, 2026
ed796af
feat(maintain-context): constrain durable project language
adamwhp Aug 15, 2026
2b9349e
feat(modular-design): clarify passive structural constraints
adamwhp Aug 15, 2026
53cc1e8
feat(tdd): enforce durable behavioural cycles
adamwhp Aug 15, 2026
1c878ae
feat(verify-change): distinguish stale evidence
adamwhp Aug 15, 2026
2108b57
feat(code-review): formalize independent inspection
adamwhp Aug 15, 2026
ce1cf4d
feat(maintain-agents): keep repository guidance lean
adamwhp Aug 15, 2026
b2a3563
feat(define-product): establish breadth-first product foundations
adamwhp Aug 15, 2026
a68a3d1
feat(define-architecture): add technical foundation workflow
adamwhp Aug 15, 2026
7dbd378
feat(specify-feature): add observable feature definition
adamwhp Aug 16, 2026
0d48dd8
feat(design-feature): add buildable solution design
adamwhp Aug 16, 2026
349dc9c
feat(implement): enforce reviewed verified candidates
adamwhp Aug 16, 2026
d23c2cb
feat(create-tickets): add vertical tracker decomposition
adamwhp Aug 16, 2026
f2f9f60
feat(commit): preserve unrelated git state
adamwhp Aug 16, 2026
d7db26c
feat(pull-request): publish complete branch outcomes
adamwhp Aug 16, 2026
0a34937
feat(review-pull-request): add external change assessment
adamwhp Aug 16, 2026
8727898
feat(triage-work): add evidence-led intake routing
adamwhp Aug 16, 2026
b2ac485
feat(plan-iteration): add feasible native work selection
adamwhp Aug 16, 2026
fb45647
feat(refine-backlog): add evidence-led backlog health
adamwhp Aug 16, 2026
673a784
feat(debug): require causal reviewed repairs
adamwhp Aug 16, 2026
21a6a4b
feat(review-architecture): simplify evidence-led assessments
adamwhp Aug 16, 2026
09448cb
feat(respond-to-incident): add mitigation-first command
adamwhp Aug 16, 2026
fc54d77
feat(review-incident): add blameless evidence-led learning
adamwhp Aug 16, 2026
9a85fb1
feat(plan-retirement): add risk-led capability exit plans
adamwhp Aug 16, 2026
0290ac7
refactor(skills): remove obsolete context elicitation
adamwhp Aug 16, 2026
95c2d76
docs: document the software delivery lifecycle
adamwhp Aug 16, 2026
53b4d02
fix(create-tickets): accept retirement authorities
adamwhp Aug 16, 2026
5ba3bfe
fix(skills): close suite acceptance gaps
adamwhp Aug 16, 2026
88c573e
feat(write-skill): replace authoring workflow
adamwhp Aug 17, 2026
19e6fa2
Partly complete skills
adamwhp Aug 18, 2026
06cf2c6
feat(elicit): strengthen information gathering
adamwhp Aug 18, 2026
d6247f3
feat(skills): delegate commit and pull request workflows
adamwhp Aug 18, 2026
39f39ad
feat(research): delegate trusted report investigation
adamwhp Aug 18, 2026
6b65637
feat(skills): split context maintenance and route elicitation
adamwhp Aug 18, 2026
fbdd5a0
chore(skills): remove prototype skill
adamwhp Aug 18, 2026
57b0eeb
feat(skills): finalize maintain-agents
adamwhp Aug 18, 2026
1695b84
docs(agents): clarify composed skill boundaries
adamwhp Aug 19, 2026
aed22be
feat(skills): consolidate product and architecture definition
adamwhp Aug 19, 2026
f8aeda9
feat(skills): consolidate feature specification workflow
adamwhp Aug 20, 2026
8b44da3
chore(skills): remove tracker workflow skills
adamwhp Aug 20, 2026
11b49c8
feat(skills): finalize create-tickets
adamwhp Aug 20, 2026
0c5459f
refactor(skills): redefine implementation skill suite
adamwhp Aug 21, 2026
5fdd2af
docs(agents): constrain skill bundle content
adamwhp Aug 21, 2026
4af63e9
feat(modular-design): deepen structural design guidance
adamwhp Aug 21, 2026
29faed9
feat(tdd): finalize test-driven development guidance
adamwhp Aug 21, 2026
adc4cee
feat(quality-harnesses): teach risk-led quality selection
adamwhp Aug 21, 2026
e83e43b
refactor(skills): remove incident workflows and invocation metadata
adamwhp Aug 21, 2026
7c0f0c9
feat(code-review): refine independent review contract
adamwhp Aug 21, 2026
0da6eb0
feat(implement): refine implementation workflow
adamwhp Aug 21, 2026
f883f35
feat(review-pull-request): finalize read-only review workflow
adamwhp Aug 21, 2026
7cbeab5
chore(skills): remove plan-retirement skill
adamwhp Aug 21, 2026
9c69942
feat(review-architecture): finalize whole-codebase review
adamwhp Aug 21, 2026
0a16623
feat(skills): finalize systematic debugging
adamwhp Aug 21, 2026
85b1109
refactor(skills): require explicit invocation
adamwhp Aug 21, 2026
99af05f
docs: document skill entry points
adamwhp Aug 21, 2026
087ead8
docs: remove unused terminology
adamwhp Aug 21, 2026
8aad172
chore(skills): remove completed plans
adamwhp Aug 21, 2026
271bbe3
style(markdown): align tables with oxfmt
adamwhp Aug 21, 2026
6b8fb30
refactor(maintain-ubiquitous-language): update resolved terms inline
adamwhp Aug 22, 2026
f6415f0
refactor(skills): remove quality harnesses
adamwhp Aug 22, 2026
8f098f9
feat(skills): add code complexity measurement
adamwhp Aug 23, 2026
04f432a
refactor(skills): clarify contracts and context
adamwhp Aug 23, 2026
1a7b617
refactor(maintain-ubiquitous-language): rename context to glossary
adamwhp Aug 23, 2026
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion .oxfmtrc.json
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
"$schema": "./node_modules/oxfmt/configuration_schema.json",
"ignorePatterns": ["node_modules/**", ".opencode/node_modules/**"],
"printWidth": 80,
"proseWrap": "never",
"proseWrap": "preserve",
"tabWidth": 4,
"singleQuote": true,
"sortImports": {
Expand Down
8 changes: 6 additions & 2 deletions AGENTS.md
Original file line number Diff line number Diff line change
@@ -1,3 +1,7 @@
- When the user corrects you with a reusable, global rule, ask if they want it added to `AGENTS.md`.
- When authoring or reviewing a skill, justify its behaviour using only context available to the agent at runtime.
- When a user correction establishes a reusable repository-wide rule, ask whether to invoke `$maintain-agents` to add it.
- When authoring or reviewing a skill, include only runtime instructions and knowledge for its agent; exclude authoring rationale, evaluation history, and composition discussion, and justify behaviour only with runtime-available context.
- When eliciting within a software project, invoke `$elicit-with-context` instead of `$elicit`; use `$elicit` directly only for non-software work or as part of `$elicit-with-context`.
- When composing skills, rely on each invoked skill's public contract; do not restate or steer its internal logic.
- Default to a user-agent workflow; introduce team roles, named accountable people, approval bodies, or third-party coordination only when the user or project evidence requires them.
- Assume the execution harness supports fresh agents; do not add fallback handling for their unavailability.
- After implementing changes run `bun run checks` before handoff.
135 changes: 0 additions & 135 deletions CONTEXT.md

This file was deleted.

39 changes: 39 additions & 0 deletions GLOSSARY.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,39 @@
# Propulsion Glossary

## Language

### Skill model

**Predictability**: The degree to which a skill makes the agent follow the same process on every run, without requiring the same output.

**User-invoked skill**: A skill the user selects explicitly. This is Propulsion's default invocation mode.

**Model-invoked skill**: A skill the agent may select autonomously or invoke from another skill. Use this exception only when autonomous discovery would naturally help during ordinary coding work often enough to earn its permanent context cost.

**Invocation policy**: Client-specific metadata within a skill bundle that controls whether an agent may select that skill implicitly. The same intent may require different policy fields in different clients.

### Skill types

**Skill type**: The authority category recorded in a skill's frontmatter as `metadata.type`. Propulsion uses exactly `performer`, `router`, `teaching`, or `utility`.

**Performer skill**: Owns one bounded requested outcome end to end, including its actions, artefacts, verification, handoff, and stopping boundary.

**Router skill**: Resolves the inputs needed to coordinate other skills and returns their results. It contains only routing unique to that outcome; called skills remain authoritative and the router neither repeats nor overrides their context.

**Teaching skill**: Supplies knowledge and decision rules inside a caller's workflow. It does not own an independent operation, mutation, artefact, verification, handoff, or stopping decision.

**Utility skill**: Owns one bounded reusable operation and returns its result, whether the operation runs in the calling agent, a fresh agent, or a deterministic tool.

### Software quality

**Deterministic measurement**: A reproducible tool-produced observation of one named property of a fixed software candidate, with its scope, components, configuration, limitations, and comparison basis preserved. An attention threshold requires appraisal but is not by itself a defect, an overall quality score, or permission to weaken behaviour or design.

### Skill benchmarking

**Authentic workflow arm**: One isolated benchmark run that gives a skill set the same implementation task, starting repository, model, tools, permissions, and environment as the other arms while allowing it to load and follow its own required skills and workflow.

**Adaptive single-run comparison**: One fresh run per benchmark arm, followed by another run only for an arm whose result is tied, failed, or surprising enough that run-to-run variation could materially change the conclusion.

**Skill-context tokens**: The exact tokenizer count of a benchmark arm's skill discovery metadata and every skill or reference file loaded during the run, reported separately from whole-run model usage.

**Runtime token usage**: The execution harness's complete input, cached-input, output, and reasoning-output token counts for one benchmark arm, excluding the independent quality evaluator.
Loading